ACADEMIC SENATE: SAN DIEGO DIVISION, 0002 UCSD, LA JOLLA, CA 92093-0002

(858) 534-3640 FAX (858) 534-4528

October 28, 2020 PROFESSOR RAJESH GUPTA Halicioğlu Data Science Institute PROFESSOR SORIN LERNER, Chair Department of Computer Science and Engineering SUBJECT: Proposed Master of Data Science (online) At its October 12, 2020 meeting, the Graduate Council approved the proposal to establish a new self-supporting graduate professional degree program of study leading to a Master of Data Science (online). The Graduate Council will forward the proposal for placement on an upcoming Representative Assembly agenda. The proposed Regulation for the Master of Data Science (online) will be reviewed by the Committee on Rules and Jurisdiction. As part of its approval, the Graduate Council is also approving an exception to the general rules for transfer credit, as stated in the General Catalog, to allow the online MDS program to accept transfer credit for up to 16 units. Courses must be taken prior to matriculation at UC San Diego. Please note that proposers may not accept applications to the program or admit students until systemwide review of the proposal is complete and the UC Office of the President has issued a final outcome. In addition, please consult with Dean of Undergraduate Education John Moore on WSCUC requirements for review of the proposed distance education program.

Sincerely, Lynn Russell, Chair Graduate Council

cc: M. Allen J. Antony S. Constable R. Continetti B. Cowan T. Javidi C. Lyons J. Moore J. Morgan K. Ng R. Rodriguez D. Salmon Y. Wollman

May 22, 2020 TO: Lynn M. Russell Chair, Graduate Council, Academic Senate, UC San Diego. Attention: Lori Hullings By email: lhullings@ucsd.edu FROM: Rajesh K. Gupta Director, Halıcıoğlu Data Science Institute (HDSI) RE: Revised Proposal for a new program on Online Master of Data Science (OMDS) Dear Graduate Council: On behalf of the Halıcıoğlu Data Science Institute (HDSI) it is my pleasure to send you a revised proposal for a new self-supporting online professional graduate degree program for the Master of Data Science degree. This proposal was originally submitted for academic senate review on October 23, 2018 leading to considerable debate on the broader topic of online degree programs. Over the past year, the senate has demonstrated support for online programs and provided clear guidelines. Based on the feedback and guidance received through both academic and administrative review processes, we have revised and updated the proposal. The program origins are in a collaboration across multiple departments by incorporating existing online “micromasters” courses into an integrated program in collaboration with the Computer Science and Engineering department. In keeping with the transdisciplinary nature of the subject, the proposal draws upon faculty from a diverse group of faculty and researchers from CSE, ECE, Cognitive Sciences, Music, SIO and Health Sciences to create a curriculum of 9 courses and one capstone course. The capstone course, like our capstone courses in the Data Science undergraduate and JSOE MAS programs is likely to engage faculty and researchers from all areas of campus. The program requires development of new online courses as detailed in the proposal. Our resource planning is accordingly provisioned to ensure successful launch of this program. We fully expect new additional elective courses to be created that are not yet listed in the proposal, particularly by more than a dozen faculty who will be joining CSE and HDSI in the months to come, particularly new faculty recruits in the areas of Machine Learning, Data Mining, Artificial Intelligence, Fairness, Trust and Accountability in Data Science. As a hub for data science at UCSD, HDSI is well positioned and ready

HALICIOĞLU DATA SCIENCE INSTITUTE 9500 GILMAN DRIVE LA JOLLA, CALIFORNIA 92093-0404

to ensure such enrichment of this program through necessary administrative support, communications, outreach as well as financial support for the development of new courses. The online nature of the program has rightly led to a close scrutiny of the program proposal by all parties that has delayed the planned launch by more than year. Fortunately, increased scrutiny by the campus has also led to clear guidelines and expectations to ensure quality and effectiveness of the program to meet an important educational need that has been long felt both internally at UC San Diego as well as by our potential students, and partners who wish to engage with the university meaningfully to put to practice advances in data sciences. HDSI is also proposing in-residence state supported graduate degree programs that are broadly accessible to talent from various domains where data science plays a crucial role. The OMDS program and its courses provide us with effective screening and on-boarding tools to attract a diverse talent to strengthen our graduate student population as well as strengthen our data and compute infrastructure of teaching data science subjects online or in the classroom. My sincere thanks to the faculty involved in preparing this proposal, for their perseverance to create new courses and to Sorin Lerner for his commitment to online education and to CSE chair Dean Tullsen in making this proposal an exemplar of departmental cooperation with HDSI. With Sorin Lerner as the incoming chair of CSE, I feel confident that the joint program is well-positioned to build upon this cooperation for an effective and successful launch of an online graduate program in data science and look forward to your support for the program. Thank you and please do feel free to contact me if you need more information or if I can answer any questions. Sincerely,

Rajesh K. Gupta

CC: Sorin Lerner, Chair-elect, Computer Science and Engineering Dean Tullsen, Chair of Computer Science and Engineering Albert Pisano, Dean of Jacobs School of Engineering Robert Continetti, Senior AVC, Academic Affairs Jennifer Morgan, MSO/HDSI Encl: Response to Graduate Council and Revised OMDS Proposal “Clean Copy” OMDS Proposal.

May 22nd, 2020

TO:

Lynn Russel

Chair Graduate Council, Academic Senate, UC San Diego

FR:

Rajesh K. Gupta, Director, Halıcıoğlu Data Science Institute

Dean Tullsen, Chair, Computer Science and Engineering

RE: Proposal for a Masters of Data Science (MDS)

Dear Graduate Council:

Thank you for the additional guidance that you have provided for online degrees. Attached is a response

to the questions posed in the guidelines, along with an updated proposal.

Sincerely,

Rajesh K. Gupta Dean Tullsen

CC:

Albert Pisano, Dean of Jacobs School of Engineering

Robert Continetti, Senior AVC, Academic Affairs

Sorin Lerner, Chair-Elect, Computer Science and Engineering

Encl: Response to Graduate Council questions, updated MDS proposal

HALICIOĞLU DATA SCIENCE INSTITUTE 9500 GILMAN DRIVE LA JOLLA, CALIFORNIA 92093-0404

A proposal for a Self-Supporting Professional Graduate

Degree Program in Data Science for the Master of Data

Science (online) degree

Developed by:

Sorin Lerner, CSE, Vice-Chair of Grad Programs, and Chair-Elect

Leo Porter, CSE, Chair of CSE MS Committee

Justin Eldridge, HDSI, Chair of HDSI Online Education Committee

Jingo Shang, HDSI and CSE

Contact Persons:

Rajesh K. Gupta

Director of the Halıcıoğlu Data Science Institute

rgupta@ucsd.edu

Dean Tullsen

Chair of Computer Science and Engineering

tullsen@eng.ucsd.edu

Version History:

October 26, 2018: Version 1.0 submitted to UCSD Graduate Council

December 3, 2018: Version 1.1 submitted to UCSD Graduate Council

April 3, 2019: Version 1.2 submitted to UCSD Graduate Council

May 22, 2020: Version 1.3 submitted to UCSD Graduate Council

A proposal for a Self-Supporting Professional Graduate Degree Program in Data Science for the Master of Data Science (online) degree

Executive Summary: Data Science is a burgeoning field of study that combines concepts from Statistics, Computer Science, and Applications where data is forefront and center. These applications are drawn from wide-ranging areas in both academia and industry, including Engineering, Physical Sciences, Social Sciences, Health & Life Sciences, and Arts & Humanities. The Halıcıoğlu Data Science Institute (HDSI), in coordination with the Computer Science Department (CSE), proposes to offer a Master of Data Science (online) to working professionals who are looking to expand their skill set into this new and exciting area of study. The goal of the program is to teach students the skills required to be successful at performing data-driven tasks. The program will be fully online, with the material delivered through high-quality videos, interactive code environments, discussion forum prompts, and formative quizzes. Students will be graded on assignments, projects and exams (through online e-proctoring). Instructors and TAs will hold online office hours and will answer questions on online forums. Students registered in the program will get a UCSD degree. We expect in the steady state to have about 300 - 600 students.

1

Contents 1. Introduction 4

1.1. Aims and Objectives of the Program 4

1.2. Historical Development of the Field 5

1.3. Timetable for development of the program 7

1.4. Relationships of the Program to Existing Programs on Campus 7

1.5. Interrelationship of the program with other UC institutions 10

1.6. Department or group which will administer the program 11

1.7. Plan for evaluation of the program 11

2. Program 13

2.1. Undergraduate preparation for admission 13

2.2. Foreign language 16

2.3. Program of Study 16

2.3.1. Specific fields of emphasis 16

2.3.2. Plan 16

2.3.3. Unit Requirements 16

2.3.4. Required and Recommended Courses 16

2.3.5. Description of capstone element 19

2.3.6. Academic Integrity 23

2.3.7. Ensuring Teaching Excellence and Overall Quality of the Program 24

2.4. Field examinations 26

2.5. Qualifying examinations 26

2.6. Thesis and/or dissertation 26

2.7. Final examination 26

2.8. Explanation of special requirements over and above Graduate Division minimum requirements. 26

2.9. Relationship of Master’s and Doctor’s programs 26

2.10. Special preparation for careers in teaching 26

2.11. Sample program 26

2.12. Normative time from matriculation to degree 27

2

3. Projected Need 27

3.1. Student demand for the program 27

3.2. Opportunities for placement of graduates 27

3.3. Importance to the discipline 28

3.4. Ways in which the program will meet the needs of society 28

3.5. Relationship of the program to research/professional-interests of the faculty 28

3.6. Program Differentiation 28

4. Faculty 29

5. Courses 30

5.1. Course delivery 30

5.2. Foundations (3 courses) 32

5.3. Core (3 courses) 33

5.4. Electives (Pick any 3 courses) 34

5.5. Capstone (1 course) 35

6. Resource requirements 35

6.1. FTE faculty 35

6.2. Staff 37

6.3. Space and other capital facilities 37

6.4. Computing costs 37

6.5. Teaching Assistants 37

6.6. Cost of course development 38

6.7. Revenue to cover expenses 38

6.8. Sample Budget 38

6.9. Academic Unit Profit 38

7. Graduate Student Support 39

7.1. Financial Accessibility 39

7.2. Diversity, Equity and Inclusion 41

8. Governance 41

9. Changes in Senate regulations 42

3

1. Introduction

1.1. Aims and Objectives of the Program

Data Science is a burgeoning field of study that combines concepts from Statistics, Computer Science, and Applications where data is forefront and center. These applications are drawn from wide-ranging areas in both academia and industry, including Engineering, Physical Sciences, Social Sciences, Health & Life Sciences, and Arts & Humanities. Data Science as a subject concerns itself with data, its interactions with humans, and its impact on society. These include processes and their models that generate data, method and tools that enable us to store, analyze, understand and visualize data, and interaction of data systems with humans and physical systems.

Our aim is to provide an online Professional Master of Data Science to working professionals who want to expand their skills into this new and exciting area of study. The learning goal of the program is to teach students the skills required to be successful at performing data-driven tasks. This includes the ability to: (1) collect raw data from various sources and convert this raw data into a curated form amenable to algorithmic analysis (2) understand machine learning algorithms and how to run them on large data sets; (3) interpret the results of these algorithms and iteratively drill down into the data, and perform more analysis, to answer questions about the data.

The program is designed to be online so that we are able to reach a broad geographic population, and provide educational experience to a community that until now has been underserved. The rising cost of higher education, along with the economic challenges faced in residential education leave behind a large population of students; many of them are our graduates from years or decades ago, who need to keep up with changing technological realities. Many of these students cannot attend residential education at any price due to career constraints or family obligations. Further, emerging areas, such as Data Science, represent a leading edge of technological advances that simply cannot be taught by community colleges or our own extension programs. Given the demand for our existing online MicroMasters courses in Data Science (see below), we anticipate there will be a large demand for this online Master degree. See Section 3.1 for a more detailed discussion of demand for the program.

We expect most students admitted to the program to have an undergraduate degree in a field that provides a good mathematical foundation, for example Computer Science, Mathematics, Engineering, Physical Sciences, Quantitative Social Sciences, Computational Life Sciences or Computational Health Sciences. However, we also expect some students to have an undergraduate degree in a field where mathematical foundations might not be at the core of the curriculum, for examples Arts & Humanities (as we will describe later, our Capstone course will offer a Computational Music project that combines machine learning with Music). This casts a wide net, which is almost a necessity in Data Science since the data always come from a particular domain, and domain knowledge is important. Moreover, many online learners are interested in a career change and we anticipate many will have learned prerequisite skills

4

outside traditional university degrees (e.g., at their present employment and through online courses). To ensure that admitted students are able to succeed in the program, we will allow students to demonstrate their readiness to enroll by taking the first four courses of the program as online MOOCs for a small fee. This concept, which is formalized as a type of edX certificate called a MicroMasters, is a new and unique way of reaching (and preparing) students, which will be explained in full detail in Section 2.1.

The Master of Data Science will be priced at a point that is comparable to other online professional Masters degrees, for example those offered by Georgia Tech, University of Michigan, and the University of Illinois, Urbana Champaign.

The formal name of the program, as it appears in the transcript, will be “Master of Data Science (online)”. The use of the word “online” is meant to indicate the modality of teaching, in the same way that transcripts at UC campuses have a special designation for online courses (at UCSD online courses in the transcript have an “R” at the end, for example 210R -- other campuses use similar designations). The use of the word online is not meant to signal a lower quality or rigor program. The quality and rigor of the online program will be at the same level as in-person programs. In the remainder of this document, we will use the following three designations for the program, depending on the English sentence: online Master of Data Science; Master of Data Science (online); or in some cases the three-letter registrar code MDS, which will also be used for courses in the program, for example MDS 210R.

1.2. Historical Development of the Field

Data Science is an emerging area driven largely by the need of organizations to make sense of the large scale of diverse data now made possible through advances in sensing, computing and storage technologies. Traditionally, large data sets were understood through the lens of statistics, a field devoted to understanding data in a mathematical framework. To do so, statisticians often start with precise mathematical models of processes that generated the data, models that are validated post-facto through tools and techniques devised by statisticians. Advances in computing enabled the notion of a “runtime” or timeliness association with data. While data science as a field continues to take shape driven by algorithmics and computing system advances, a prevailing school of thought places this field at the intersection of statistics and computer science with a strong role of multiple application domains. We briefly review 1

these underpinnings as they inform us of the steps we should take to devise an effective graduate program in Data Science.

Historically data analysis has been a domain of Statistics, a branch of Mathematics that deals with the analysis and interpretation of data. While exploration of statistics has been around for centuries, Statistics started becoming a more substantive field of inquiry in the 18th century, with the use of probability and calculus to understand data. Many universities now have entire Statistics departments. At UC San Diego, statistics faculty are part of the Mathematics department and closely affiliated (in different roles) with the HDSI. As Statistics matured with

1 David Blei, Padhraic Smyth, “Science and Data Science,” PNAS August 2017, http://www.pnas.org/content/pnas/early/2017/08/04/1702076114.full.pdf

5

strong mathematical foundations and practical methods, computer science emerged as providing new sets of means to analyze data.

Computer Science is concerned with the study of computation. It emerged as a discipline in the 1950s and 1960s with the advent of physical computers capable of general computation. With the advent of Computer Science, the field of Statistics became more computational, leveraging computers to perform statistical analysis. At the same time, early computer scientists were immediately struck with the idea of using machines to “think”, leading to the subfield of Artificial Intelligence (AI). AI was originally concerned with using logic to perform human-like thinking and reasoning. Over time, these early logic-based approaches gave way to data-driven approaches now known as machine learning: such approaches can learn from data using concepts like probabilities and/or neural networks. Machine learning, which was originally explored in the 1980s and 1990s, has seen a rebirth in the past decade, in part because of the tremendous success of classification and optimization techniques and their implementations such as stochastic gradient descent by neural networks. This success has been enabled in part because of computational innovations like GPUs and large clusters of computers. Recent advances in Machine Learning have created stunning results such as AlphaGo, the first program to defeat a Go world champion (a longstanding challenge in the Artificial Intelligence community). Such successes have raised the possibility of a much more intelligent and automated world of systems that sense, actuate and adapt.

The final catalyst for the emergence of Data Science is the advent of vast quantities of data, something often dubbed “Big Data”. With the price of storage going down, we are in an era where large amounts of data are routinely collected and stored for analysis: companies collect data about their customers and inventory; scientists gather increasingly large amounts of data from ubiquitous sensors; economists have increasing numbers of economic indicators, in addition to live-streaming market data; universities collect many kinds of data about their students’ preparation and performance.

With all this data coming together, along with the maturation of the fields of machine learning and computing, we have come to a juncture where Data Science is emerging. Data Science overlaps with other fields, including Mathematics, Statistics, Computer Science, Machine Learning, and many application domains; but the central tenet of Data Science, which makes it unique from these other fields, is that the data itself takes the forefront: Data Science is a field that makes data itself the primary entity of interest, and through this lens investigates all of the surrounding issues, including, for example how to create, store, curate, analyze, understand, visualize and disseminate data.

Because Data Science is a new field that UCSD is investing in, UCSD has already created several new degree programs in Data Science, including an in-person Master of Advanced Study in Data Science Engineering, and an undergraduate program in Data Science (more details about these in Section 1.5). UCSD has also recently launched the Halıcıoğlu Data Science Institute (HDSI), a new academic unit capable of recruiting and housing faculty members to advance Data Science as a field. The proposed online Master of Data Science will

6

be one of the first graduate offerings enabled by HDSI and will add to UCSD’s growing academic offerings in Data Science that engage multiple academic units and research centers.

1.3. Timetable for development of the program

Proposal submitted for UCSD approval June 2020

Proposal submitted to CCGA October 2020

CCGA Approval (UC-level approval) December 2020

WSCUC Approval (Accreditation agency) March 2021

Applications Accepted March-June 2021

Admissions of first Cohort August 2021

Program Offered October 2021

The above table shows a timeline of the program approval process. In the timeline, we will accept applications after CCGA approval, and concurrently with WSCUC approval (which WSCUC allows). It is important to realize that since this is a professional Master program, it can still be successful even if it does not follow the more traditional December deadline for graduate school applications. We plan to have two admission cycles per year. Initial enrollments will start at about 100 students. Over time, we expect enrollments will be somewhere in the range of 300 - 600 students per year. These targets will be evaluated continuously as we look at the quality of the students, the quality of our delivery, and the bandwidth we have for teaching. For comparison, the fully-online MS in Computer Science at Georgia Tech has about 6,000 students per year; the Berkeley Master of Information and Data Science (all online except for a short 3-4 day on-site immersion) has 600 students per year. We already had more than 280,000 students enroll in the first course of our MicroMasters of Data Science since it began in Fall 2017. Of these students, about 3,700 were verified learners (paid students). This strong demand for our MicroMasters classes is encouraging and (we believe) an indicator that there will be strong demand for our Data Science Master program.

1.4. Relationships of the Program to Existing Programs on Campus

Existing MS Degrees: There is a Data Science and Machine learning concentration in the on-site UCSD Electrical and Computer Engineering (ECE) MS program. There is also a concentration by the same name in the ECE PhD program, although for the purposes of this document, the concentration in the MS program is more relevant. The Computer Science and Engineering (CSE) department also has a large on-site MS program in Computer Science, and many of those students choose to specialize in Artificial Intelligence. The student population in these existing on-site programs (ECE MS and CSE MS) consists mostly of foreign students who study in the US either to gain access to the US job market, or to get involved in research and jump into a PhD program. Because our proposed online Master of Data Science would not

7

provide those opportunities (access to job market or research), we believe it would have a minimal impact on the demand for the CSE and ECE MS programs.

MAS in Data Science Engineering: The CSE department, in collaboration with ECE and the San Diego Supercomputer Center, offers an on-site professional Master of Advanced Study (MAS) in Data Science Engineering. This is a small program of about 30-40 students a year, with a relatively high price-point of about $39,000. The MAS started in Fall 2014.

The proposed online Master of Data Science (hereto often referred to as MDS) is an outgrowth of the MAS in Data Science Engineering. Indeed, a few years ago several instructors from the MAS program started building online edX versions of four MAS classes, and these four classes were formalized as an edX Data Science MicroMasters. This early investment in online education for Data Science has now placed us in a perfect position to create a fully online Master of Data Science program: we already have the first four classes of the online program built and fine-tuned; and the edX MicroMasters certificate provides us with the ideal pipeline for recruiting students into the program. Building on this MicroMasters is a great opportunity for UCSD.

While there is some overlap in the material between the on-site MAS and the online Master of Data Science, our proposed online program is different from the MAS in five ways:

(1) Curriculum: The curriculum for the online Master of Data Science (MDS) will be more interdisciplinary than the MAS in Data Science Engineering. The faculty who are engaging with the MDS program come from fields as disparate as engineering, music, cognitive science, and anthropology. At launch the capstone project will include project options from Music, Oceanography, and Computer Vision. Over time, we expect to offer additional capstones and electives from various disciplines, for example Engineering, Health Sciences, Social Sciences, Physical Sciences, Life Sciences, and Arts & Humanities. The MAS in Data Science Engineering, as indicated by the word “engineering”, focuses more on the engineering aspects of data science than the MDS. This difference in focus manifest itself in several ways including: the MAS has two required classes on storage systems for big data, whereas the MDS has only one; the electives in the MAS focus more on issues related to low-level system building, system performance, and analysis, whereas the electives in the MDS program cover a wider spectrum of data science, including courses like Putting Ideas into Production, and Interaction Design. Another difference in the curriculum concerns the capstone. The MAS in Data Science Engineering has a two quarter customized capstone where a group of students work together to analyze previously unexplored data from a company, whereas the MDS offers a one-quarter capstone where students choose from a small number of pre-designed projects.

(2) Price point: Our proposed online degree costs $22,000 and is therefore much less expensive than the on-site MAS.

(3) Program size: The number of students in the online program will be much larger than the MAS program.

8

(4) Level of instructional support: The on-site MAS degree provides more one-on-one instructor support, along with a personalized capstone, in which small groups of students work closely with the professor to define and explore a custom project; in contrast the capstone for our online Master of Data Science will be a pre-designed project.

(5) Geographic flexibility: Most attendees in the MAS degree are from Southern California, whereas the online Master of Data Science will attract a much broader audience.

To make sure students understand the difference between the MAS and the MDS programs, we will design a web page that compares the two programs, and also explains the trade-off between in-person and online education.

We believe that the online Master of Data Science can coexist with the on-site MAS, offering two different options from which students may choose: a less expensive fully online degree without a customized capstone; and a more expensive degree with more mentoring and a customized capstone. Still, there is a small chance that the online program will consume some of the student pool of the MAS to the point that the MAS program will not be viable anymore. If this were to happen, we would have to sunset the MAS program, which would be unfortunate. However, there are two important points to note about this scenario. First, it would allow us to consolidate our efforts into what the students would appear to want more: a cheaper online option without a customized capstone. Second, and more importantly, if this scenario were to happen, we strongly believe it would happen even if UCSD didn’t start an online Master of Data Science. As described in Section 3.6, other universities are starting online professional master programs in Data Science. If we assume that a cheaper online program will cannibalize the MAS program to the point of not being viable, then those other programs would similarly lead to the demise of the MAS program, even without an online offering from UCSD.

Undergraduate Program: UCSD also started offering an undergraduate major and undergraduate minor in Data Science in Fall 2017. The undergraduate program targets a different audience than our online professional master program, and as such the two programs will not compete with each other. In fact, the two programs are complementary: alumni from the undergraduate program may be good candidates for the online Master of Data Science, either immediately, or after several years in industry. The undergraduate major already has 609 majors, just a year after being launched. The popularity of the undergraduate major not only shows how much interest there is in Data Science, but also provides evidence that there will likely be a large UCSD Data Science alumni pool that can be leveraged for recruitment into the online Master of Data Science.

Impact on enrollments and class sizes of state-supported programs at UCSD: As discussed above, the enrollments and class sizes in MS and undergraduate programs at UCSD will not be affected by the new online program. In fact, the online program will have a positive impact on the course offerings in state-supported programs. Indeed, the material developed for the online program (and possibly even some of the classes themselves) can be repurposed for state-supported students. Some of the developed projects for example could also be repurposed for our undergraduate major in Data Science, or for our MS in CSE program. There

9

is also a huge demand for graduate-level Data Science classes across the UCSD campus: many PhD students in application fields want to get some Data Science basics to analyze their data. The online Master of Data Science will allow us to leverage our efforts in online education to help these students, either by allowing them to directly enroll in our MDS classes, or by offering a similar class through a state-supported program.

1.5. Interrelationship of the program with other UC institutions

The Berkeley School of Information offers a Master of Information and Data Science. This is also an online self-funded professional master degree. The tuition for the entire program is about $63,000. All the courses are offered online, and each online section is about 30 students. The program has a total of about 600 students. Students are required to attend at least one, 3–4 day immersion on the UC Berkeley campus. During the immersion, students meet with classmates and professors in person, participate in workshops, and have opportunities to network with data science professionals.

There are five differences between our proposed online Master of Data Science and the Berkeley online Master of Information and Data Science: (1) price point: the Berkeley degree costs about $63,000 whereas our proposed degree costs $22,000 (2) class size: the Berkeley program has small class sizes of 30 students for one instructor; the UCSD program will have larger class sizes, on the order of hundreds of students per instructor -- but with a lot of student support (3) immersion: the Berkeley program has an immersion requirement, as described above; the immersion requirement clearly provides value to the student, but also makes it harder for non-local students to participate in the program (and especially for international students who need to obtain a visa) (4) MicroMasters: The UCSD program makes use of a novel and unique mechanism, the MicroMasters certificate, for attracting students from diverse backgrounds, while ensuring that students have the preparation needed; details on how this functions are described in Section 2.1 (5) curriculum: The UCSD program aims to be interdisciplinary in nature; at launch the capstone project will include project options from Music, Oceanography, and Computer Vision; and over time, we expect to offer additional capstones and electives from various disciplines, for example Engineering, Health & Life Sciences, Social Sciences, Physical Sciences, and Arts & Humanities.

We see the Berkeley program and our program as two distinct points in the design space. Both programs are online, but they have different price points, along with different educational tradeoffs. It is also important to realize that our program and the Berkeley program (and in fact all online degrees) will be selective in their admissions. Ultimately, each program will only capture a small finite portion of the global online educational market for Data Science professional degrees. So there is plenty of room for many different programs in this space. We believe that both our online program and the Berkeley program can co-exist successfully.

10

1.6. Department or group which will administer the program

The program will be offered by the Halıcıoğlu Data Science Institute (HDSI) and the Department of Computer Science and Engineering. Teaching in the program will be done by faculty from many units on campus. A more detailed governance structure is outlined in Section 8.

1.7. Plan for evaluation of the program

The program will be formally evaluated like all other UCSD graduate programs, consistent with Senate regulations, every 8 years. This formal evaluation will include an external review and UCSD graduate council oversight.

Learning Outcomes and Metrics. Before launch, the program will define learning outcomes for the entire program and for each course, and will clearly define how the learning outcomes will be assessed in each course, and how learning outcomes for each course align with the learning outcomes for the entire program. All of this is in fact required for WSCUC approval, which will come after CCGA approval. The program will work closely with the UCSD Director of Digital Learning and Teaching + Learning Commons to develop the learning outcomes, the assessment plans, and how the collected information will be used to improve the program.

In addition to the above evaluation of learning outcomes, the program will also use standard metrics that are part of all program reviews, including: program size, applicant pool size, admission rate (% of applicants offered admission), yield (% of admitted students who accept), GPA, graduation rate, time-to-degree, proportion of woman and URM in the program (i.e.: in the applicant pool, admitted pool, new student pool, overall student body, and faculty), and post-graduation placement of students.

The one metric that we want to be careful with is time-to-degree. While we will definitely collect time-to-degree, it is also important to realize that in a professional Master program where students might be balancing several different obligations, time-to-degree is not always the best indicator of success. Our ultimate goal is to provide a quality program that works for the schedules of our students, which means that time-to-degree is less important than for in-person full-time students.

Another metric that we want to track carefully is completion rate. Completion rate will be one of the many metrics we will use to assess the long-term viability of the program. Informed by the metrics of comparable programs, like Georgia Tech’s online Masters program, and other online Masters offered by EdX, we expect a completion rate for the program to be in the 80-85% range. A drop below 80% would be a warning sign. A drop below 60% would be a red flag, in which case we will consider drastic remedial measures, and if those do not work, we will work with UCSD Graduate Council to forge a path forward, possibly even cancelling the program. We will measure completion rate as the ratio of the number of students who complete the degree within 5 years to the number of students who start the degree enrolled at UCSD. This 5-year completion rate will take into account all students, even those who take a longer time to graduate. This is especially important because many of the students will be working

11

professionals who might only be taking one class per quarter, and also juggling other commitments.

The program will also work with the office of the Chief Information Officer and with EdX to determine what data can be collected about online students, and how this data can be used to help increase student success. The UCSD Chief Information Officer, Vince Kellen, was hired in 2016, and has a demonstrated track record at prior institutions of using innovative applications of information technology in teaching, learning and student success. We look forward to working with Vince Kellen and his office to ensure the highest rates of student success in our online program.

Monitoring and Evaluation. As mentioned above, the program will officially be evaluated by the Senate like all other programs, on a 7-8 year period. However, since the start of the program will be the most critical time, we plan to undergo a formal evaluation of our program yearly for the first five years with reviews that include: (1) completion rates for individual classes; (2) completion rates for the degree program; (3) average time to degree; (4) demographic information; (5) each statistic in 1-3 broken down by demographic category; (6) course, TA, and professor response surveys, including student comments and peer instructor reviews; (7) an evaluation from Digital Learning in the Teaching+Learning Commons; and (8) financial data.

The above review cycles will be supplemented by a continuous internal evaluation process, which, broadly speaking, will include student feedback & surveys, teaching evaluations, alumni feedback, and industry feedback. Based on our experience with the launching of our Data Science and Engineering MAS and the recent 4-fold growth of the Computer Science & Engineering MS, we believe that continuous internal evaluation, followed up by periodic and deliberate evolution is the key to success. More specifically, as described in full detail in Section 2.3.7, we will appoint a Teaching Excellence Coordinator for the program, a UCSD professor with expertise in online teaching who will work toward making the program of the highest quality. This person will continually evaluate student and instructor performance to understand how to optimize program delivery and student success. We will intervene when problems arise. Interventions could include: working with the professor of specific classes to improve course delivery and assessments; adding additional TA support for classes where students are struggling; providing additional one-on-one TA and instructor support for students who are having problems.

Diversity. We will pay particular attention to evaluating how our program performs in terms of diversity, equity and inclusion. However, defining success for Diversity, Equity, and Inclusion is difficult in an online program and particularly for an online program in a field where in-person programs struggle with higher failure rates and lower retention of students from underrepresented groups (women and URM in computing). However, we expect online programs to be capable of reaching a different demographic than our in-person classes, including students from lower socioeconomic status, working professionals, and older learners. We will continuously examine demographic data and student outcomes, by demographic group, to look for areas of concern. Where there are discrepancies in outcomes by group (e.g., URMs fail at a higher rate than non URMs), we will compare those outcomes with our in-person

12

courses and published outcomes from other online programs to see if the MDS is doing particularly better or worse by comparison. Should problems be identified, we will work with the Teaching and Learning Commons, as well as leverage the large body of research on inclusive teaching from the education community, to improve these measures.

To understand how our numbers for diversity, equity and inclusion compare to the national norms, we will compare our numbers against both in-person Master programs and prominent online Master programs. For comparison, the Online Master of Computer Science at Georgia Tech, one of the premier online Master of Computer Science, publishes its numbers. These 2

numbers show that at Georgia Tech’s Online Master of Computer Science Women are 17.3% of the student population, and under-represented minorities are 11.7%. For comparison, in the UCSD CSE in-person MS in Computer Science, Women are 22.3% of the student population, and under-represented minorities are 4.1%. The CSE in-person MS in Computer Science has higher proportions of women than Georgia Tech’s online program, but a lower proportion of under-represented minorities. Our goal will be to do better than Georgia Tech’s online program and also better than our in-person programs. This aggressive goal will be facilitated by a variety of efforts, including outreach efforts, holistic review of applications, training of the admissions committee on inclusive admissions, and diversity fellowships.

If we find that our diversity, equity and inclusion numbers are significantly worse than either the CSE in-person program or Georgia Tech’s online program, it will be time for aggressive measures to rectify the situation, including additional outreach, additional efforts on inclusive admissions, and additional funds for diversity fellowships. If the numbers do not improve with these aggressive measures, we will work with UCSD Graduate Council to forge a path forward.

2. Program

2.1. Undergraduate preparation for admission

We expect most students admitted to the program to have an undergraduate degree in a field that provides a good mathematical foundation, for example Computer Science, Mathematics, Engineering, Physical Sciences, Quantitative Social Sciences, Computational Life Sciences or Computational Health Sciences. However, applications of Data Science can also be found in fields where undergraduate degrees might not cover mathematical foundations, for examples Arts & Humanities. Because we want our program to offer as many interdisciplinary opportunities as possible (for example, one of our capstone projects will be in Music), we also would like to admit some students from these disciplines. This casts a wide net, which is a necessity in an interdisciplinary field like Data Science, where the data always come from a particular domain, and knowledge of that domain can be crucial. The wide variety of student backgrounds will be a strength of the program, but at the same time we must carefully manage the admissions process in the context of this wide net to make sure all students have the required skills to succeed. We will work with the Graduate Division at UCSD to implement

2 https://www.omscs.gatech.edu/prospective-students/numbers

13

appropriate structures to avoid bias in admission decisions, including faculty training and the use of rubrics and metrics.

Data Science MicroMasters. One mechanism we will leverage to ensure student success is our existing Data Science MicroMasters. A MicroMasters is a new type of edX certificate that consists of a four-course sequence offered online through the edX platform. The four courses in a MicroMasters must have e-proctored exams that account for a majority of the grade. UCSD has already developed a Data Science MicroMasters, which consists of the first four courses of our proposed online Master of Data Science. In other words, we are offering the first four courses of the MDS, online, for anybody to take. There is no fee to audit these four MicroMaster classes. However, to get the actual MicroMasters certificate (which involves taking the e-proctored exams), students must sign up as Verified Learners and pay a fee of $350 per course. It is important to realize that anybody can sign up as a Verified Learner to take the MicroMasters, even students who are not matriculated at UCSD.

The four classes in the MicroMasters provide a good foundation for the rest of the MDS program. As such, doing well in the MicroMasters is a good indicator that someone can do well in the rest of the program. Performance on the MicroMasters can then be used as one (of many) criteria in admission decisions. Some students may have such a strong prior academic record that they do not need to complete the MicroMasters to demonstrate their skills prior to admission. However, for other students, most notably students from non-traditional backgrounds, the MicroMasters is a unique opportunity: it allows such students to take their time in learning the basic material, possibly even taking additional remedial online classes to fill in their knowledge gaps. This ultimately enables these students to demonstrate that they can in fact learn the material and do well in the program, whereas in a more traditional admission setting, such students might not even be strongly considered for admission. This mechanism we hope will allow us to evaluate students who come from various disciplinary backgrounds, including Engineering, Health & Life Sciences, Social Sciences, Physical Sciences, and Arts & Humanities.

While the MicroMasters allows UCSD to evaluate students before admitting them, the reverse is also true: the MicroMasters allows prospective students to evaluate UCSD before they commit to the entire cost (in terms of money and time) of a Master degree. For a relatively small fee, they can see how instructors teach the classes, what kind of production value we put into our material, what kind of assignments we give, and what kind of exams we use.

We envision a model where students have the option of taking the MicroMasters before matriculating but are not required to do so. If a student takes the MicroMasters before matriculating at UCSD, and the student is later admitted to MDS, credit from the MicroMasters will be transferred to MDS, meaning that the student will not be required to retake the first four classes of MDS. If a student is directly admitted into MDS without having first completed the MicroMasters, the student will simply complete the MicroMasters courses after matriculating.

Students who complete the MicroMasters courses after matriculating will enroll in the courses through UCSD. The MicroMasters courses for students enrolled at UCSD will be the same as

14

the MicroMasters courses for the verified learners who are paying on edX. However, the edX students and the UCSD students will be in separate “sections” (distinguished primarily by being in separate online forums). Furthemore, the students taking the course as UCSD students will also get more exclusive access to TAs and instructors (similar to what UCSD students get in the rest of the program). We envision that the fees for the MicroMasters courses will be $1,400 USD, regardless of whether the courses are taken on edX or at UCSD. The fee for the six remaining courses at UCSD will be $20,600. Thus, regardless of what path students take through the program (all UCSD vs MicroMaster transfer), students pay $1,400 for the first four courses and $20,600 for the six remaining courses. This structure allows students to gradually commit to the Masters program, while understanding what they are getting into. The gradual commitment occurs in three stages.

1. Students can start by auditing the courses on edX for free. At this point, they can see the lecture material and quizzes, but not the full assignments and final.

2. Students who like the lecture delivery and feel they follow the lectures can enroll in the MicroMasters, which consists of the first four courses in the online Master of Data Science program. There is no admission process to the MicroMasters. The MicroMasters allows students to try out 40% of the online Master of Data Science program with a low barrier to entry. The low barrier to entry is not just in terms of price ($1,400, which is 6% of the total price of the program for 40% of the courses), but also in terms of admission process for the MicroMasters: students can try the program without having to apply (i.e.: without having to write a statement of purposes, without having to get letters of reference, and without having the possible barrier of being rejected by the admissions committee). If the student completes the MicroMasters, they get an edX certificate, regardless of whether they continue in the Master program. The MicroMaster certificate itself has value: on the very first offering of the MicroMasters, 82 students have already paid and completed the full MicroMaster certificate, without even having the option to transfer the credit into a Master program. A recent study also found that completing a 3

MOOC data science program that costs less than $500 led to an average increase in salary of $8,230 and an increase in the likelihood of job mobility. This again shows that a MicroMasters certificate has value even for students who end up not continuing in the Master program. Finally, students pay and complete the MicroMasters as they go, providing additional gradual commitment: each course costs $350 and each time a course is completed, the student gets a certificate of completion for the course. The full MicroMaster certificate is awarded at the end of the four cases.

3. Students who do well in the MicroMasters and want to continue will likely want to apply to the online Master of Data Science program. Good performance in the MicroMasters can be a good indicator of future success in the Master program, which helps students

3 https://papers.ssrn.com/sol3/papers.cfm?abstract_id=3260695

15

decide if they want to commit the additional time and money, and also helps the Master admissions committee make decisions.

In summary, we believe allowing all students, without admission, to take 40% of the program for 6% of the program tuition provides a good path for gradual commitment to the program. This is beneficial for students because it allows students to see what online education is, and whether they would be successful in this setting, without committing to the entire program. Furthermore, it also allows the admissions committee to make more informed decisions.

Admissions requirements to MDS are equivalent to those for the Master of Science:

● Bachelor’s degree in a relevant field ● UCSD Data Science MicroMasters recommended but not required ● Undergraduate GPA of at least 3.0 on 4.0 scale ● Three letters of recommendation ● TOEFL or TSE scores (international applicants only)

2.2. Foreign language

Foreign language is not required for this degree.

2.3. Program of Study

2.3.1. Specific fields of emphasis

The pipeline of activities that Data Scientists perform is as follows: (1) collect raw data (2) prepare the raw data, which includes curating, cleaning, formatting, and preprocessing (3) store the curated, prepared data set into some form of large storage system (4) use mathematical foundations to guide the analysis of the data (5) analyze the data, using a variety of tools (6) visualize the results. This pipeline of 6 stages outlines the major fields of study. Each course will cover several stages of this pipeline, to various extents. After we describe the courses in 2.3.4, we will show, in Figure 1, a matrix of how each course covers each of the 6 above stages.

2.3.2. Plan

The program will follow Master Plan II, capstone project. The capstone project will be described in more detail in Section 2.3.5 (as required in the proposal format).

2.3.3. Unit Requirements

The program will require 10 quarter-length courses at 4 units each, for a total of 40 units.

2.3.4. Required and Recommended Courses

The classes will be organized in four categories: Foundations (take all 3 courses), Core (take all 3 courses), Electives (pick any 3 courses) and Capstone (1 course). We list below the starting set of courses in each category, along with a one-sentence description for each course. More detailed descriptions can be found in Section 5 (as required in the proposal format).

16

Foundations (take all 3 courses)

The foundation courses provide the basic background needed in the remainder of the program.

MDS 200R: Python for Data Science. This course introduces students to several important tools that are needed in Data Science, including: Python, Jupyter Notebooks, Pandas, NumPy, Matplotlib, Scikit-learn, and the NLTK Data.

MDS 210R: Probability and Statistics in Data Science using Python. This course covers the foundations of probability and statistics needed for Data Science.

MDS 220R: Machine Learning Fundamentals. This course covers a variety of basic supervised and unsupervised learning algorithms, and the theory behind those algorithms.

Core (take all 3 courses)

The core courses cover the central topics of the program.

MDS 230R: Big Data Analytics Using Spark. This course covers practical techniques for doing data analysis on large amounts of data, using tools such as MapReduce, Hadoop and Spark.

MDS 240R: Data Mining on the Web. This course covers the application of several Machine Learning and Data Mining techniques to a variety of applications, including: text mining, playlist prediction, suggestion for smart reply, learning visual clothing style, and online advertising.

MDS 250R: Data Management for Analytics. This course covers how to store and manage large amounts of data, with an eye toward applications in Data Science.

Electives (Pick any 3 courses)

Students will be able to customize their experience in the program by taking 3 electives. We list here several electives that the program will offer. At launch we plan to have at least 3 electives; within one year after launch, we plan to have at least the 5 electives listed below. Over time, we will add additional electives to provide students with more options.

MDS 260R: Advanced Unsupervised Learning. This course covers advanced techniques for unsupervised machine learning, including dimensionality reduction, k-means, principal component analysis, topic modeling and deep unsupervised models.

MDS 261R: From Data to Products. This course teaches students how to build an entire data-processing pipeline that can be used in production.

MDS 262R: Data Visualization. This course covers techniques for creating effective visualizations to explore trends, identify relationships, test hypotheses, communicate findings and gain insight about data.

MDS 263R: Data Preprocessing. This courses covers techniques for taking raw data (for example collected directly from web pages) and converting it into a clean data set that machine learning techniques can work on.

17

MDS 264R: Interaction Design. This course introduces fundamental methods and principles for designing, implementing, and evaluating user interfaces for exploring data.

18

Capstone (1 course)

MDS 298R: Capstone Project in Data Science. This course consists of a quarter-long project. Students will pick one project out of several available options, each project from a different domain. At launch we expect to have projects in: Music, Oceanography, and Computer Vision. Over time, we expect to add additional capstone projects from various disciplines, for example Engineering, Health & Life Sciences, Social Sciences, Physical Sciences, and Arts & Humanities.

Overview Matrix

Figure 1 shows an overview matrix of the courses in the program.

Recall from Section 2.3.1 the pipeline of activities that Data Scientists perform: (1) collect raw data (2) prepare the raw data into a curated data set (3) store the curated data set in a storage system (4) use mathematical foundations to guide the analysis of the data (5) analyse the data (6) visualize the results.

Figure 1 shows the starting set of courses in the program, and the amount of material in each course for the 6 stages of the pipeline. For the main stages that a course focuses on, we also list the main topics covered for that stage of the pipeline. The height of the green bar in each stage indicates the relative amount of material focused on that stage of the pipeline.

2.3.5. Description of capstone element

The capstone project will consist of one large quarter-long project. Students will choose from several projects that have been carefully designed and curated. Here are some examples of possible projects in various domains:

1. A Music Project, which would consist of training a machine learning model on existing music to create a model that can generate new music in a given style. Shlomo Dubnov, Professor in the Music Department at UCSD, has an on-site course on precisely this topic, Music 206: Experimental Studies Seminar: Deep and Shallow.

2. An Oceanography project, which would consist of developing predictive and analytical models with data from the famed Scripps Institution of Oceanography.

3. A Computer Vision project, which would use machine learning to identify pedestrians and obstacles in the setting of, for instance, self-driving cars.

The capstone course will consist of two parts, which will happen simultaneously. The first part will introduce some background material in the application domain. The second part will apply techniques learned throughout the program to a project in the given domain. The project will cover all topics in the program, including preparing, storing, analyzing, and visualizing data.

The project will be defined and designed ahead of time. This allows instructors to create projects that are well thought out from a pedagogical point of view. These pre-designed projects will need to be approved by the program director and the Teaching Excellence Coordinator, to ensure that they meet the course standards.

19

20

By default, students will NOT be able to design their own projects. While individualized projects provide some benefit, including fostering creativity, it can be hard to ensure a consistently good selection of projects for hundreds of students.

For each project, there will be one instructor with expertise in the domain of the project (which we call a “domain instructor”), and several TAs with expertise in that field. There will also be one instructor who oversees the entire class, and participates in final grading (as described below). We call this instructor the “common instructor”.

The project will be individual. Students will submit two artifacts: (1) their code and (2) a project report. The grading will be done as follows:

1. Code: a rigorous automated test suite will be used to evaluate the student’s code. 2. Report: The report will be evaluated by a TA with expertise in the field, using a rubric

established and agreed upon by all instructors, including the common instructor for the entire capstone class (who will ensure that the rubric is fair across areas). The TAs doing the grading will be overseen by the domain instructors and the common instructor, providing oversight to ensure consistent grading. In particular, each domain instructor will oversee the grading for the project in their domain. This will be done in two ways: (a) the TA will consult with the domain instructor on tricky grading cases (b) the domain instructor will randomly sample about 10% of the projects, and carefully look over the grading that the TA has done for those 10% of the projects. Any problems or inconsistencies discovered will have to be addressed by the TA not only in the 10%, but also across all other exams. This process of random spot-checking will be done in parallel with TA grading, to catch problems early on. This is a common practice to ensure proper oversight of TAs when grading a large number of exams. Finally, the common instructor will do quality assurance in two ways. First, the common instructor will ensure that the domain instructors are properly overseeing their TAs in grading (i.e. that the domain instructors follow their duties as described in the prior paragraph) Second, the common instructor will also do random spot-checking of about 10% of projects, across domains. The common instructor will report and make sure that any discovered inconsistencies are addressed.

While the CCGA guidelines (Appendix I in the 2016 Handbook) mention at least two reviewers for each report, we believe that in our capstone a single TA evaluating each report is enough to ensure high-quality evaluations for four reasons: (1) the capstone consists of pre-designed projects with well described requirements, which reduces the variation in the content of the reports (2) the reports will be evaluated using a careful and consistent rubric (3) the TA evaluating the reports will have expertise in the field, and will be overseen by an instructor (4) a large part of the capstone grade will come from an evaluation of the code, which will be done independently of the report evaluation.

Justification for Pre-designed Projects. As stated in the CCGA handbook, “capstone projects should be synthetic, tying together two or more areas of specific content that would typically be the subject of a class or a sequence of classes.” With this in mind, our goal in the capstone

21

project is to provide a culminating project in which students apply many aspects of the material they learned throughout the program in a single project.

We see a tradeoff between pre-designed capstone projects and customized capstone projects. Pre-designed capstone projects can be carefully designed and fine-tuned to achieve learning goals, and to cover a wide gamut of concepts that span the entirety of the program. However, they rarely provide an open-ended exploration of a topic. Customized projects do provide an open-ended exploration of a new topic, but on the other hand, can sometimes fail to meet the learning goals of the capstone course. This is because each customized project, by its very nature, is being explored for the first time during the course, and sometime unexpected obstacles arise that prevent students from exploring the entire gamut of concepts. As a result, to make sure customized capstone projects meet learning goals, instructors must provide a lot of mentoring to the students. With 300 to 500 students doing capstone projects each year, it becomes all the more difficult to provide customized projects, while also ensuring that all projects achieve the learning goals.

To inform our decision, we also looked at what other online programs do. The summary is that both pre-designed and customized capstone projects exist in online data science programs. Typically, higher-cost programs offer customized projects, whereas lower-cost programs offer pre-designed projects. For example, the Berkeley Master of Information and Data Sciences, with a tuition of $63,000, has a customized capstone. At lower price point of $19,200, the online Master of Computer Science in Data Science at UIUC (a top 10 school in Computer Science) offers pre-designed projects in their capstone course. 4

In designing our program, we considered the tradeoffs between pre-designed vs customized capstone, the target size of our program, our tuition point, and what other online programs do. In the end, we opted for a default of pre-designed projects. While the projects are pre-designed, they will still meet the CCGA requirement that the capstone project ties together concepts from the entire program. Furthermore, we will also offer students several such pre-designed projects, to cover of range of possible interests from students.

Finally, as we gain experience in delivering our capstone, we will also allow a small select set of students, based on grades and interest, to do a custom project. We envision having at most 10 students a year doing this. Students doing custom projects can work individually or in groups. Such students would enroll in an honors version of the capstone course, and would get more individual mentoring through skype/zoom. The honors version of the course would be marked as such in the transcript with a different code, for example “MDS 299R Honors Capstone Project in Data Science”. Students would apply to participate in the honors version of the project by proposing a project in a statement of purpose. The common instructor of the class will evaluate these statements of purpose and coordinate with possible instructors and researchers to determine if there is the possibility of accommodating the project. The common instructor will also make sure all custom projects are consistent with course standards. For these custom

4 https://cs.illinois.edu/sites/default/files/docs/syllabi/CS598_CloudComputingCapstone.pdf

22

projects, students will submit their code and a written report. Students in the custom projects will also be required to do a presentation online. For grading, there will be no automated grading. The report and presentation will be evaluated by two faculty, one being a domain instructor (including the possibility of a faculty advisor), and one being the common instructor, who will have no relationship with the student.

As with all aspects of the program, we will closely monitor the capstone projects to make sure they meet the highest standard of quality, and they also meet student needs. If we find problems, we will make sure to address them immediately.

2.3.6. Academic Integrity

Even though a section on Academic Integrity is not required in the proposal format, we add it here because of its importance in any educational endeavor, and particularly in an online one. To uphold the highest standards of academic integrity, we will take the following steps:

● E-proctoring for exams: We will use a combination of e-proctoring services provided by Software Secure and ProctorU. E-proctoring records a full video of the person taking the exam through the person’s web camera, along with a full capture of the screen and all keyboard and mouse events. Staff from Software Secure or ProctorU will check the IDs of students, and monitor the live feeds of all exam takers. E-proctoring provides a high-level of oversight, since the feeds are constantly monitored, and all the feeds are recorded for later inspection.

● Plagiarism Detection Tools for Assignments/Projects: We will use plagiarism detection tools (e.g., MOSS, TurnItIn). Such plagiarism detection tools have been used heavily in existing CSE courses.

● Following best practices. We will follow best practices as recommended by the UCSD Academic Integrity Office. These practices include: discussing academic integrity in the syllabus and in the first lecture; providing students with a link that explains the appropriate academic integrity policies and processes; clearly stating the grading policy for violations of academic integrity; carefully documenting all allegations of academic integrity violations; having students sign an Academic Honesty Statement each time they submit an assignment or exam; and running plagiarism detection software on all assignments.

● Academic Integrity in the Capstone: We will pay special attention to academic integrity for the Capstone project. First, the program will change the projects over time. Each year, small changes will be made so that solutions from previous years cannot be copied directly. Furthermore, every few years, a major project overhaul will take place. Second, as stated above, we will use plagiarism detection tools. We will make heavy use of a particular tool called Moss, which has become a standard for plagiarism detection in computer science. The Computer Science and Engineering department has a lot of experience using Moss for large in-person classes (classes with over 200 students, both at the undergraduate level and at the MS level). Moss takes a set of assignment submissions and compares each submission which all other submissions. It provides a report of similarity for each pair of submissions. Moss is based on a sophisticated

23

algorithm that would not be fooled by simple changes made to an assignment. Put differently, if a student copies someone else’s assignment and makes only minor changes, Moss will detect this. To take into account submissions from prior years, we will simply add the submissions from the capstone project of prior years. We have done precisely this for large in-person classes, and it works well. As mentioned earlier too, the capstones will change over time, and so prior solutions should not work year by year.

● Reporting Allegations of Academic Integrity Violations: Consistent with the UCSD academic integrity policy, we will immediately report all allegations of academic integrity violations for UCSD students to the UCSD Academic Integrity Office, and follow their process for resolution. If we discover academic integrity violations for non-UCSD students in MicroMasters classes, we will follow the academic integrity policy and process of edX, and report these allegations to edX.

● Academic Integrity Coordinator: Because we want to enforce the highest level of academic integrity, an Academic Integrity Coordinator (AIC) will be appointed. The AIC will be a UCSD professor whose main role will be to ensure that all courses follow all the bullet items in this list. If needed, the efforts of the AIC will be supplemented with staff or tutor support, for example, to organize/manage allegations of academic integrity violation, and to run plagiarism detection software.

The UCSD department of Computer Science and Engineering has a long tradition of taking academic integrity seriously. UCSD gives out yearly awards to celebrate faculty, staff, students and departments/units who are champions of academic integrity at UCSD. CSE has received several of these awards, including two Academic Integrity Faculty Awards (2013 and 2015) and the 2017 Academic Integrity Department Award.

Continuing on this tradition, all of the academic units involved in offering the online Master of Data Science will uphold the highest standards of academic integrity in this new program. If needed, the efforts of the AIC will be supplemented with staff or tutor support, for example, to organize/manage allegations of academic integrity violation, and to run plagiarism detection software.

2.3.7. Ensuring Teaching Excellence and Overall Quality of the Program

To ensure teaching excellence, we will appoint a Teaching Excellence Coordinator (TEC). The TEC will be a member of the academic senate whose main responsibility will be to ensure that the program’s educational quality is of the highest caliber. The TEC will be a professor with extensive knowledge and expertise in the field of online education.

The CSE department has several very highly rated MOOCs, that are known to be among the leading MOOCs in their areas:

● The “Python for Data Science” course on the edX platform taught by Leo Porter (Associate Teaching Professor in CSE) and Ilkay Altintas (Director at SDSC, Lecturer in

24

CSE, and HDSI Fellow). This is the first course in the proposed online program, and is on the “Best Courses of 2017” list at classcentral.com 5

● The “Computer Graphics” course on the edX platform taught by Ravi Ramamoorthi (Professor in CSE). Ravi Ramamoorthi was a finalist in 2016 and 2017 for the edX prize for exceptional contributions in online teaching and learning.

● The “Bioinformatics Specialization” on Coursera taught by Pavel Pevzner (Professor in CSE)

● The “Interactive Design Specialization” on Coursera taught by Scott Klemmer (Professor in CSE and Cognitive Sciences)

● The “Object Oriented Java Programming: Data Structures and Beyond Specialization” on Coursera taught by Mia Minnes (Associate Teaching Professor in CSE), Christine Alvarado (Teaching Professor in CSE) and Leo Porter (Associate Teaching Professor in CSE).

The instructors of the above MOOCs could be perfect candidates for the position of TEC.

The TEC will work with the UCSD Director of Digital Learning, and the UCSD Teaching + Learning Commons to set instructional guidelines and best-practices for the program. The TEC will work with instructors to apply these best practices in building and preparing the material for the classes in the program. Indeed, with the growing demand for online programs, there has been considerable research on student retention rates and strategies to improve the online course experience. Several factors lead to attrition but a leading one is the misconception learners have about the workload and general expectations. The Digital Learning group, led by Karen Flammer, is developing an online orientation program that describes the rigors and demands of each course, time expectations and strategies for weekly time management. Additionally, this orientation program will have a module on how to navigate the edX and edX Studio platform and use the “Course Progress Dashboard” to track performance and progress through each course. Verified learners in the MicroMasters program and students accepted to the MDS program will be required to take this online orientation. Given that MicroMasters learners are from diverse backgrounds, social factors can also play a role in determining student retention. The literature on online education (a survey of which can be found here: https://journals.sagepub.com/doi/full/10.1177/2158244015621777) suggests that enhancing social interactions in an online program can help with student morale and student retention. We already find that learners in the MicroMasters courses exchange personal emails, join each other’s LinkedIn, discuss their current employment and even make plans to meet at future conferences. The MDS program will explore further mechanisms to foster a sense of belonging to the UCSD community, and to encourage networking and collaboration. Examples would include: course/program website where students, faculty and instructional assistants can upload a photo and a short bio; live TA sessions; live instructor office hours; orientations to help students manage their online presence.

5 https://www.class-central.com/report/best-free-online-courses-2017/

25

The TEC will also monitor all the metrics discussed in Section 1.7, and immediately take appropriate action when issues arise.

2.4. Field examinations

No field examinations are required

2.5. Qualifying examinations

No Qualifying examinations are required.

2.6. Thesis and/or dissertation

No thesis or dissertation is required.

2.7. Final examination

No final examination is required. Evaluation of the capstone project will serve as the final examination for the program.

2.8. Explanation of special requirements over and above Graduate Division minimum requirements.

There are no special requirements over and above Graduate Division requirements.

2.9. Relationship of Master’s and Doctor’s programs

This program will result in a supplemental Master’s degree with no relationship to a doctoral program.

2.10. Special preparation for careers in teaching

Teaching is not a component of this degree.

2.11. Sample program

For a sample program, we will show the common pathway that we expect to see: a student taking the four MicroMasters courses online as a non-matriculated student, and then taking the remaining 6 courses as a matriculated student.

Year 1 (as non-matriculated student): DSE 200, DSE 210, DSE 220, DSE 230 as part of MicroMasters (the courses are scheduled in such a way that they can be taken in a year)

Year 2 (as matriculated student): 2 courses per quarter for a total of 6 courses

Fall: DSE 240 (Data Mining on the Web), DSE 250 (Data Management for Analytics)

Winter: DSE 260 (Advanced Unsupervised Learning), DSE 261 (From Data to Products)

Spring: DSE 264 (Interaction Design), DSE 298 (Capstone)

26

2.12. Normative time from matriculation to degree

We expect one common pathway to be: a student takes one year for the 4 MicroMasters as a non-matriculated student, and then one year for the remaining 6 courses as a matriculated student. This will be a total of 2 years, one year as a non-matriculated student, and one year as a matriculated student. Of course, there will be a lot of variation depending on: (1) what other obligations students are trying to balance with their studies (2) at what point during the MicroMasters the student chooses to apply and matriculate -- some students might apply and matriculate before completing the entire (or any of the) MicroMasters.

3. Projected Need

3.1. Student demand for the program

IBM in May 2017 predicted that demand for data scientists will soar 28% by 2020. 6

Bloomberg in May 2018 labeled “Data Scientist” the hottest job in America. 7

A comprehensive 549 page report, released in June 2018 by SNS Telecom & IT (a market intelligence & consulting company), estimates that Big Data investments will account for over $65 billion in 2018 alone, and that these investments will grow at a annual rate of approximately 14% over the next three years. 8

In June 2018, Google’s chief economist claimed that the world needs more data scientists. 9

A coherent picture emerges from all these articles: the need for Data Scientists is growing at a very fast pace, and there aren’t enough Data Scientists to fill the demand. One driver of this growth is that companies across all sectors are trying to leverage the data they collect by distilling it into actionable information. Furthermore, there is no established data science workforce, and there are very few undergraduate programs specific to Data Science.

In this setting, our online program allows professionals who already have an undergraduate degree to gain skills in Data Science, and make themselves more marketable, both within their own company, or to other companies.

We have seen very strong demand for our Data Science MicroMasters, which is a good predictor for our online Master of Data Science. Our MicroMasters just started in 2017 and we already have almost 4,000 verified learners (paying students) enrolled.

3.2. Opportunities for placement of graduates

We expect students who graduate from our program to get promotions and/or salary increases within their company, or to transition to a new job at a different company.

6 https://bit.ly/2NYoDdn 7 https://bloom.bg/2ItFjrj 8 http://www.snstelecom.com/bigdata 9 https://bit.ly/2kOIS0V

27

One comparison data point is the online Master of Information and Data Science from the Berkeley Information School. The program states on their website: 10

“30 percent of recent graduates reported receiving a promotion. 58 percent of recent graduates reported getting a new job. 72 percent of recent graduates reported receiving a salary increase.”

Furthermore, UCSD has a good reputation for producing strong technical graduates. UCSD was ranked 9th nationally in the most graduates hired by Silicon Valley, only 2nd after Berkeley in the UC system. We believe that the additional skills and credentials provided by a professional 11

Master of Data Science from UCSD will help students develop their careers.

3.3. Importance to the discipline

A professional degree in Data Science is important for several reasons.

First, there is an immediate need for Data Scientists in industry, but very few undergraduate programs in this area. As such, there is a great need for an affordable professional degree that can imbue some data science expertise to existing members of the workforce.

Second, as previously explained, Data Science is an interdisciplinary field that draws from Statistics, Computer Science, and Application domains where data analysis is needed. The interdisciplinary nature of Data Science creates the perfect stage for a professional Master degree of the kind we propose, with a MicroMasters that allows students from various backgrounds the opportunity to earn a place in the program.

3.4. Ways in which the program will meet the needs of society

As explained previously, Data Scientists are highly sought after, showing a societal need for individuals with this professional competency. Furthermore, Data Science and Machine Learning are having an impact on many aspects of society, including e-commerce, financial industries, technology companies, health care, and academia. There are few aspects of society that will not be affected by Data Science.

3.5. Relationship of the program to research/professional-interests of the faculty

As can be seen from Section 4 that lists faculty, and in the Appendix of faculty CVs, the program directly aligns with the research interests of professors participating in the program.

3.6. Program Differentiation

In Sections 1.4 and 1.5 we have already discussed the related programs at UCSD and in the UC system. We now discuss related programs outside of the UC.

There are many on-site Master degrees nationally in areas related to Data Science. These programs are often offered through a unit like Computer Science, Statistics, Electrical Engineering, or an Information School, with a specialization/concentration in Data Science or

10 https://datascience.berkeley.edu/ 11 https://bit.ly/2q37s

28

Artificial Intelligence. One key way in which our program differentiates itself from these programs in that we offer a fully online program. The differences between these on-site Master degrees and our online degree are similar to the differences described in Section 1.4 between the existing on-site UCSD MS degrees and our online degrees. We refer the reader to Section 1.4, under the heading “Existing MS Degrees”, for a discussion of the differences.

In terms of online offerings, there are only a handful of online Master programs that are related to Data Science. We discuss here the most prominent players in this area:

● The Georgia Institute of Technology (Georgia Tech) was a pioneer in this field. Georgia Tech was the first highly ranked school (a top 10 Computer Science school) to create a large, online, low-cost Master program. Their program is an MS in Computer Science, 12

and costs about $7,000 for the entire degree. There are a total of 6,000 students enrolled in the program.

● University of Illinois offers a Master of Computer Science in Data Science. The 13

program is a fully-online professional master program, which takes 12 to 36 months to complete depending on the pace, and costs $19,200.

● John Hopkins offers an MS in Data Science that can be taken online, or partly online and partly on-site. 14

● University of Michigan has offered a Master of Applied Data Science since the Fall of 2019. The total cost of the 34-credit degree is approximately $31,688 for in-state students and $42,262 for out-of-state students, given this year’s estimated rates before fees. 15

The UCSD online Master of Data Science also differentiates itself from the online and in-person programs offered elsewhere in its strongly interdisciplinary approach to data science. The faculty who are engaging with the program come from fields as disparate as engineering, music, cognitive science, and anthropology (see below). At launch the capstone project will include project options from Music, Oceanography, and Computer Vision. Over time, we expect to offer additional capstones and electives from various disciplines, for example Engineering, Health Sciences, Social Sciences, Physical Sciences, Life Sciences, and Arts & Humanities.

4. Faculty There are many faculty who have shown interest in being involved in the program. We list the faculty here, categorized by expertise area. Faculty in a given area are listed alphabetically by last name, along with departmental affiliation (CSE: Computer Science and Engineering; ECE: Electrical and Computer Engineering; SDSC: San Diego Supercomputer Center; CogSci: Cognitive Science; SIO: Scripps Institution of Oceanography).

12 http://www.omscs.gatech.edu/ 13 https://online.illinois.edu/mcs-ds 14 https://ep.jhu.edu/programs-and-courses/programs/data-science 15 https://www.coursera.org/degrees/master-of-applied-data-science-umich

29

● Artificial Intelligence & Machine Learning: Ilkay Altintas (SDSC), Sanjoy Dasgupta (CSE), Yoav Freund (CSE), Julian McAuley (CSE), Alon Orlitsky (ECE)

● Basic Programming: Leo Porter (CSE) ● Storages Systems and Data Processing: Arun Kumar (CSE), Alin Deutsch (CSE),

Yannis Papakonstantinou (CSE) ● Visualization and User Interaction: Scott Klemmer (CSE & CogSci) ● Application Domains (with department and domain): Manmohan Chandraker (CSE,

Vision), Shlomo Dubnov (Music, Generative models for music), Peter Gerstoft (SIO, Machine learning for physical applications), Jade d'Alpoim Guedes (SIO & Anthropology, Computational models in Anthropology), Pavel Pevzner (CSE, Bioinformatics), Hao Su (CSE, Vision)

In addition, with the recent creation of the Halıcıoğlu Data Science Institute (HDSI) as an academic unit capable of housing faculty appointments, there will be significant faculty hiring in Data Sciences at UCSD in the foreseeable future. We believe there will be no shortage of faculty who will be qualified to teach in the program.

5. Courses

5.1. Course delivery

Instruction. All courses will be delivered through a combination of the edX online platform and the edX Studio hosted at UCSD. The course content will include:

● Video lectures. Video mini-lectures (4-20 minutes) teach course content as well as provide live-coding examples which encourage students to work through their own code in parallel with the instructor. In addition, there are a series of guest lectures which introduce students to data science faculty at UCSD and SDSC.

● Readings. Students are given brief readings with either code or course content to review.

● Discussion forum prompts. Students are prompted periodically to engage in discussions regarding course content. They are given questions to help facilitate that discussion.

● Formative quizzes. Ungraded short practice quizzes allow student to get feedback on their learning periodically in the course.

● Graded quizzes. Graded quizzes are short quizzes that are not worth a large portion of the grade. Students will not be proctored during these quizzes. This mirrors a practice for in-person courses where some instructors use online graded quizzes that are not proctored.

● Projects. Some classes will have project components, which will be graded using a variety of approaches, including: (1) automated grading (2) TA/Instructor grading (3) peer grading with careful rubrics and appealable grades (note that peer grading is also a practice that is beginning to gain traction for in-person classes)

30

● Exams. Exams will be proctored using Software Secure, a leading e-proctoring system that verifies student identities and monitors students during exams. When an exam starts, Software Secure authenticates students using a photo ID through a webcam, and requires a room scan of the test environment. During the exam, Software Secure monitors and records the learner (through a webcam) and computer desktop (using desktop monitoring software) to catch any inappropriate physical or online activity. The course team can specify the rules (e.g.: no phones, no calculators, no going to other websites, etc.). Any suspicious event is reported by the Software Secure team to the instructor at UC San Diego to make a determination. Each proctored event (audio + video) is recorded and can later be replayed to look at problematic cases, thus providing a significant level of monitoring. This kind of e-proctoring has already been approved by UCSD Graduate Council for graduate R (remote) classes.

Office hours: The instructor and each TA will hold two kinds of office hours: scheduled office hours, and ad-hoc office hours.

Scheduled office hours will happen at advertised times. done either through online video chatting (using tools like Google hangouts, Google chat), or through online question forums (where the idea is that at certain assigned times, there will be someone monitoring questions non-stop and answering right away) The instructor will use feedback throughout the class to fine tune the amount of scheduled office hours, and appropriate times for these office hours (especially considering the global audience of the program, which includes issues with time zones)

The second kind of office hours will be ad-hoc office hours. These will consist of instructors and TAs logging into online forums periodically to answer questions that have been posted. For example, each TA might be asked to spend 30 minutes a day answering questions throughout the day (for example, go online 6 times a day and answer questions for 5 minutes each time).

Training TAs: In addition to taking our department’s required CSE 599 (Teaching Methods in Computer Science) course, the instructor will be responsible for training the TAs to make sure they are effective in the setting of an online class. This will include making them familiar with the online platforms, understanding how to answer questions online, and how to hold online office hours. Even through the class is online, the instructor and the TAs will meet in person once a week to manage the class (and of course will coordinate through email and other online forums more frequently)

Instructor role: The instructor will monitor forums to make sure that questions are answered appropriately. In an online setting, the instructor can also weigh in on questions that have already been answered. In fact, office hours through an online forum provide much more oversight than traditional in-person office hours. Indeed, in a traditional setting, the in-person interactions between a TA and students are not recorded and not looked at by the instructor. If a TA answers a question incorrectly (or just not as well as they should have) during in-person office hours, neither the student nor the instructor find out about it. However, in an online forum,

31

the openness and transparency provides oversight not only from the instructor but from other TAs and other students, who can also weigh in on the discussion.

By seeing all the questions that are asked, the instructor can use the online office hours and online forums to get a good sense of where student misconceptions are. The instructor can then address these misconceptions by providing additional content, or additional assignments. Furthermore, in a online setting, there is also the potential for a fine-grained assessment of student mis-understandings. Indeed, because students are progressing in the course online (which can include practice quizzes to test understanding), the instructor can look at the rate of progression of students to pinpoint where students are getting confused. For example, if the instructor notices that, say question 3 of the test quiz of the second online lecture is disproportionately answered wrong, the instructor will know that the concept covered by that question is a source of confusion.

5.2. Foundations (3 courses)

MDS 200R: Python for Data Science: This course introduces students to several important tools that are needed in Data Science, including: Python, Jupyter Notebooks, Pandas, NumPy, Matplotlib, Scikit-learn, and the NLTK Data. Students will learn to find answers within large datasets by using Python tools to import data, explore it, analyze it, learn from it, visualize it, and ultimately generate easily shareable reports. Covered topics include: Basic process of data science; Python and Jupyter Notebooks; An applied understanding of how to manipulate and analyze uncurated datasets; Basic statistical analysis and machine learning methods; How to effectively visualize results.

Status: Already developed by Ilkay Altintas (Director at SDSC and Lecturer in CSE) and Leo Porter (Associate Teaching Professor in CSE) as part of the UCSD Data Science MicroMasters; EdX website: https://goo.gl/Y6XJJo

MDS 210R: Probability and Statistics in Data Science using Python: This course covers the foundations of probability and statistics needed for Data Science. Students will learn both the mathematical theory, and get hands-on experience of applying this theory to actual data using Jupyter notebooks. Covered topics include: Random variables; Dependence; Correlation; Regression; PCA; Entropy and MDL.

Status: Already developed by Alon Orlitsky (Professor, ECE) as part of the UCSD Data Science MicroMasters; Edx website: https://goo.gl/v4X9np

MDS 220R: Machine Learning Fundamentals: This course covers a variety of supervised and unsupervised learning algorithms, and the theory behind those algorithms. Students will apply techniques from this course in programming assignments with Python and Jupyter Notebooks. Using real-world case studies, students will learn how to classify images, identify salient topics in a corpus of documents, partition people according to personality profiles, and automatically capture the semantic structure of words and use it to categorize documents. Covered topics include: Classification, regression, and conditional probability estimation; Generative and discriminative models; Linear models and extensions to nonlinearity using kernel methods;

32

Ensemble methods: boosting, bagging, random forests; Representation learning: clustering, dimensionality reduction, autoencoders, deep neural networks.

Status: Already developed by Sanjoy Dasgupta (Professor, CSE) as part of the UCSD Data Science MicroMasters; EdX website: https://goo.gl/Ta762M

5.3. Core (3 courses)

MDS 230R: Big Data Analytics Using Spark: This course covers techniques for achieving scalability in data analysis, using tools such as MapReduce, Hadoop and Spark. Students will learn what the bottlenecks are in massive parallel computations and how to use the Spark framework to minimize these bottlenecks. Students will also learn how to perform supervised and unsupervised machine learning on massive datasets using the Machine Learning Library (MLlib). Covered topics include: Programming Spark using Pyspark; Identifying the computational tradeoffs in a Spark application; Performing data loading and cleaning using Spark and Parquet; Modeling data through statistical and machine learning methods.

Status: Already developed by Yoav Freund (Professor, CSE) as part of the UCSD Data Science MicroMasters; EdX website: https://goo.gl/e76x9h

MDS 240R: Data Mining on the Web: This course covers recommender systems, data mining, and predictive analytics. Students will learn how to build models that help us understand data in order to gain insights and make predictions. The course presents the material using a variety of applications as examples, including: Text mining, Playlist prediction, Suggestion for Smart Reply, Learning Visual Clothing Style, and Online Advertising. All programming assignments are in Python. Covered topics include: regression; classification; unsupervised learning and dimensionality reduction; recommender systems; text mining; social network analysis; visualization; crawling; online advertising.

Status: Julian McAuley (Assistant Professor, CSE) wants to develop this class. It would be an online version of CSE 258A, an extremely popular class that Julian McAuley has taught many times.

MDS 250R: Data Management for Analytics: This course covers how to store and manage large amounts of data, with an eye toward applications in Data Science. The course will introduce a variety of data formats, data models, high-level query languages and programming abstractions representative of the needs of modern data analytic tasks. The list of topics includes: relational (SQL-based) database systems; a variety of hierarchical graph database systems (e.g the family of JSON database systems including MongoDB, CouchBase, SparkSQL); unrestricted graph database systems (with high-level query languages including Cypher, Gremlin, SparQL, etc.);array databases, which treat data as matrices and support the class of matrix operations fundamental to machine learning algorithms (e.g. SciDB, System R); parallel programming abstractions including Map/Reduce and its descendants developed for graph data (Think-Like-A-Vertex, Gather-Apply-Scatter, PowerGraph, Edge-Map/Vertex-Reduce).

33

Status: The faculty in the database group of the CSE department (Arun Kumar, Yannis Papakonstantinou, Alin Deutsch) are interested in developing this class.

5.4. Electives (Pick any 3 courses)

At launch the program will have 3 electives ready. Over time, the program will grow to the 5 electives listed below. Eventually the program will grow to additional electives.

MDS 260R: Advanced Unsupervised Learning: This course covers advanced techniques for unsupervised learning. Covered topics include: dimensionality reduction; k-means; principal component analysis; topic modeling; deep unsupervised models.

Status: Sanjoy Dasgupta (Professor, CSE) wants to develop this class. His previous course (MDS 220R) is already successful.

MDS 261R: From Data to Products: This course teaches students how to build an entire data-processing pipeline that can be used in production. Covered topics include: mechanisms for putting machine learning techniques into production; data cleaning; data visualization.

Status: This class was already developed in the summer of 2018 by Julian McAuley (Assistant Professor, CSE) and Ilkay Altintas (Director at SDSC and Lecturer for CSE) and is available presently on Coursera (with a 4.2 out of 5.0 rating).

MDS 262R: Data Visualization: The course covers techniques for creating effective visualizations to explore trends, identify relationships, confirm hypotheses, communicate findings and gain insight about data. This course will focus on teaching students the principles and techniques for creating visual representation from raw data. The course exercises will be based on publicly available datasets and utilize freely available tools like D3.js.

Status: This will be an online version of an existing class from the Data Science Engineering MAS, DSE 241.

MDS 263R: Data Preprocessing: This course covers techniques for taking raw data (for example collected directly from web pages) and converting it into a clean data set that machine learning techniques can work on. The goal of this course is to understand the nature of information heterogeneity, the techniques of relating information from different sources, and the machinery required for achieving the integration.

Status: The faculty in the database group of the CSE department (Arun Kumar, Yannis Papakonstantinou, Alin Deutsch) are interested in developing this class.

MDS 264R: Interaction Design: Users interact with data through user interfaces. This course introduces fundamental methods and principles for designing, implementing, and evaluating user interfaces. Students will learn how to generate design ideas, techniques for quickly prototyping them, and how to use prototypes to get feedback from other stakeholders like teammates, clients, and users. Students will apply their knowledge in a final project, which was designed with teams in Silicon Valley across multiple industries to capture real-world design

34

challenges. Covered topics include: user-centered design, rapid prototyping, experimentation, direct manipulation, cognitive principles, visual design, social software, software tools.

Status: Already developed by Scott Klemmer (Professor, CSE & CogSci) as part of a Coursera Specialization with a 4.5 out of 5.0 rating. 16

5.5. Capstone (1 course)

MDS 298R: Capstone Project in Data Science: This course consists of a quarter-long project. Students will pick one project out of several available options, each project from a different domain. At launch we expect to have projects in: Music, Oceanography, and Computer Vision. Over time, we expect to add additional capstone projects from various disciplines, for example Engineering, Health & Life Sciences, Social Sciences, Physical Sciences, and Arts & Humanities. The project will require students to apply the material they learned throughout the program to practice.

Status: There are several instructors who have shown interest in developing projects for this class, including: Shlomo Dubnov (machine learning for Music), Pavel Pevzner (Big data for bioinformatics), Hao Su (Machine learning for vision).

6. Resource requirements We describe the main costs first, then the revenue to cover these costs.

6.1. FTE faculty

We will offer this program by engaging a number of faculty from various academic units through a combination of off-load and on-load teaching. Off-load teaching is done (and paid) in addition to the regular faculty obligations. This is similar to the way faculty can get paid for consulting, which is done in addition to their regular obligations. On-load teaching is done as part of the regular teaching load, and the academic unit that is contributing the faculty effort gets reimbursed the full cost of the faculty effort.

On-load versus off-load teaching is an important issue that requires constant monitoring and adjustments as needed. As a matter of principle, we are committed to ensuring excellence in instruction, taught with senate faculty, and with the constraint of not affecting state-funded programs. The exact ratio of on-load vs off-load is difficult to predict because of teaching workload variations across participating units in the program. In some cases, offload teaching might be the only way to have senate faculty meaningfully participate in the program without affecting state-supported programs. In other cases, onload teaching would work perfectly because there is excess teaching capacity. Adjusting the ratio of on-load vs off-load will be the key to ensuring that state-supported programs are not affected. Ultimately, deciding between on-load vs off-load is a resource management issue that balances all teaching obligations across different units and different programs, while maximizing participation by senate faculty.

16 https://www.coursera.org/specializations/interaction-design

35

Having said this, we expect to have 40% of the teaching done onload from the start. We will aim for the percent of onload teaching to not go below 40%, subject to the constraint of not affecting state-funded programs. As campus continues to hire more data science faculty, we will eventually be able to do more MDS teaching onload without compromising state-funded programs. Finally, as will also be explained later, department revenue will partly be used to fund additional senate FTEs, which will raise overall teaching capacity (and, thereby, onload percentage) over time.

We provide some estimates of the money we will pay for various tasks in the case of offload teaching. These numbers are subject to change, as we learn more about the work involved in each case.

● 2 ninth salary (2 months salary) to develop an online class. ● ½ ninth salary (½ month salary) to offer an instance of the class ● ½ ninth salary (½ month salary) to do minor updates to a class ● 1 ninth salary (1 month salary) to do a major update to a class

We also anticipate to compensate the Director with a stipend of 1 ninth of salary. The Teaching Excellence Coordinator and the Academic Integrity Coordinator each receive a stipend of 1/2 ninth salary.

Faculty participation in state-supported programs. Most of the faculty in the program will be ladder rank faculty. One important consideration is the impact that the online Master of Data Science program will have on faculty participation in state-supported programs. We will make sure to adjust the on-load vs off-load balance appropriately so that we do not affect teaching in state-supported programs. UCSD has recently launched the Halıcıoğlu Data Science Institute (HDSI), a new academic unit capable of recruiting and housing faculty members to advance Data Science as a field. With the creation of HDSI, there will be significant faculty hiring in Data Sciences at UCSD in the foreseeable future. This will allow us over time to increase the amount of on-load teaching, without affecting state-supported programs.

Another consideration is advising in state-supported PhD programs: we believe the MDS program will only have minimal impact on PhD advising. PhD advising bandwidth can only be affected in the case of off-load teaching (whereas for on-load teaching, a professor simply trades one teaching obligation for another, without increasing teaching load). There are several reasons why we believe the MDS will not affect advising bandwidth. First, only a small number of faculty from any given academic unit will participate in the program (in CSE, for example, it would be about 6 out of 60 faculty FTEs). Second, in the steady state, each faculty will only spend a small amount of off-load time in the program: after the first year or two of course creation, we anticipate that any faculty doing off-load teaching will participate at the rate of ½ month/year in the program (teaching one course per year), which is a relatively small amount of faculty time (note: this money can also be thought of as being used instead of summer salary for faculty who don’t have enough funding to pay for summer salary). Furthermore, this is not a new precedent: faculty already spend time and get paid outside of their regular

36

teaching/research/service requirements (indeed, these activities, just like the teaching in MDS, must be reported in APM form 025).

Finally, to offset any potential (and small) reduction of faculty participation in state-supported PhD advising, the profits from the MDS program will be used to enhance and grow departments as a whole, including the state-supported programs. This will be a net positive, as we believe it will largely outweigh any slight reduction in advising bandwidth. Indeed, the profits will be used toward new senate faculty FTE, start-up packages, fellowships for all students (PhD, MS, MDS), diversity fellowships, diversity initiatives, and mentorship programs. Furthermore, a large portion of the expenses of the program will go to hiring TAs, most of which will be MS students in a state-funded program, thus making MS programs more enticing (MS students by default don’t get any funding). In the end, we believe all these uses of the money to support growth will actually lead to an overall growth in state-supported PhD programs, and a healthier and better funded state-supported MS program.

6.2. Staff

We will hire two staff to advise students, advertise the program, do outreach, help run admissions, and help students with career planning. We estimate this at about $200,000 / year. There will be minimal impact on staff support for state-supported programs because the MDS program will have its own staff.

6.3. Space and other capital facilities

Because the program is online, it will have a minimal impact on physical infrastructure. The only space requirement will be two offices for the staff, which will be easy to find. The online courses are created and curated in existing multi-media facilities in SDSC and in the Qualcomm Institute that the MDS faculty will have access to.

6.4. Computing costs

Students will need access to cloud computing resources to run data analyses. Based on offerings from the MicroMasters, we estimate this to be about $15/course of cloud computing infrastructure, which will be about $150 per student going through the entire program.

6.5. Teaching Assistants

For matriculated students, we will aim in the steady state to have one TA per one hundred students. However, in the first few years, we will allocate more TAs to make sure that first-time delivery issues are compensated by good TA support. In particular, we will assume one TA per 20 students for the first cohort, one TA per 50 students for the second Cohort, and then one TA per 100 students after that. The TAs will be expected to run online office hours, answer questions on forums, update assignments and grading scripts, and run plagiarism detection tools.

The Jacobs School of Engineering receives and admits a large number of MS students each year in the CSE and ECE departments. Most of these MS students study Artificial Intelligence

37

and Machine Learning. Similarly, the Cognitive Science department (in the Division of Social Science) and the Mathematics department (in the Division of Physical Sciences) have growing graduate student populations interested in Data Science. Drawing from all these departments (CSE, ECE, Cognitive Science, Mathematics), there is a large pool of potential TAs for use in the MDS program. In fact, because MS students don’t typically get research assistantships, they are often eager to get TA positions. As a result, the TA positions created by the MDS program will provide well needed support for MS students, and will make state-supported MS programs more appealing to students.

TAs for the MDS will be required to take the CSE department’s course on effective teaching: CSE 599: Teaching Methods in Computer Science. The course provides students with guidance on how to best engage students in discussions and Office Hours, design rubrics, communicate effectively with students and other instructional staff, effective pedagogies for teaching computing, how to assess learning, how to encourage and enforce Academic Integrity, and how to create an inclusive environment for a diverse student population. In addition, we will work with the Director of Online Learning to ensure best practices in online teaching are taught to TAs in the MDS program.

The cost of a TA is affected by a variety of factors, most importantly whether the department has to pay for non-resident supplemental tuition. If the department does not need to pay nonresident supplemental tuition, a TA costs $12,557.40 per quarter. If the department pays nonresident supplemental tuition, the cost is $17,591.40. Based on prior statistics, the average cost of a TA is $13,211.82 per quarter (the cost is close to the lower end because the department does not pay the nonresident supplemental tuition for TAs who are MS students).

6.6. Cost of course development

We already discussed that faculty will be paid 2 ninths salary (2 months salary) to develop a class. For online classes, there are also class production costs, which for example include video and studio production costs. The program will be granted access to the same institutional infrastructure (in the Digital Learning Hub and Information Technology Services) that the university is presently building to support the creation of online courses and degree programs that advance university priorities (see attached letter from the EVC).

6.7. Revenue to cover expenses

The costs described above will be covered by revenue from the program. The program charges $22,000 for the entire program.

This academic unit revenue will cover program expenses like TAs, Instructors and Staff. Whatever is left over from the academic unit revenue after expenses are paid will be called academic unit profit.

6.8. Sample Budget

Appendix A shows a budget for the first 5 years of the program in the format required by UCOP. This budget was developed in coordination with the UCSD Campus Budget Office. The UCSD

38

Campus Budget Office has given its final approval on this budget. The budget in Appendix A does not include one-time costs before the program is launched. In particular, it does not account for the fact that about 6 to 10 courses will have to be created, accounting for a one-time cost of about $240,000 to $400,000 (for instructor pay) before the program starts. This one-time expense and the deficit in the first year will require an initial monetary investment. This initial investment will be provided by the Halıcıoğlu Data Science Institute (HDSI) and the Computer Science Department (CSE).

6.9. Academic Unit Profit

At 500 students, the MDS program will generate an academic unit profit of about $8,500 per student. Because the data science program will be interdisciplinary, it will involve many academic units across campus. The academic unit profit will be distributed to the various academic units involved in the program, in proportion to the participation of each academic unit in the teaching and course development in the program. As a result, the profits from the program will benefit many units across campus, including units from Engineering, Physical Sciences, Social Sciences, Health & Life Sciences, and Arts & Humanities.

The academic unit profit will be used to advance the educational and research mission of the university. This includes using the profits for: funding senate FTEs, start-up packages in faculty hiring; fellowships for all students (PhD, MS, MDS); fellowships for contributions to diversity; diversity, equity and inclusion initiatives; and mentorship programs.

At 500 students in the MDS program, the UCSD campus will also receive $7,000 per student (separate from financial aid and from the above academic unit profit), which can be used to advance the mission of the entire university.

7. Graduate Student Support

7.1. Financial Accessibility

The program will provide financially accessible education in threeways.

First, the fee structure of the program itself makes the program affordable. The cost for the entire Master degree is $22,000. This price point is significantly lower than typical Master programs, thus providing an affordable path to a Master degree.

Second, 10% of the UCSD revenue (before it reaches the academic unit) will be allocated to financial aid. Some students will be eligible for financial aid of up to $5,000, based on financial need.

Third, the MicroMasters is specifically designed to help with financial risk. Indeed, as was explained in Section 2.1, the MicroMasters allows for gradual financial commitment. More specifically, the first four courses, which are part of the MicroMasters (whether done through edX or through UCSD), costs a total of $1,400 (which is paid as the student takes each course, at $400 per course). This means that students can do 40% of the program for 6% of the

39

total cost. We believe that 40% of the program is a large enough portion that students can understand what the program is about, how online education works, and whether the program is a good match for the student’s interests and needs. As a result, we believe the MicroMasters provides a great mechanism for students to understand what they are getting into with a relatively low financial risk. The remaining six courses cost a total of $20,600, or $3,433 per course. Here again, students pay as they take the classes. For example, students could take 7 courses for $8,130 ($1,400 + 3 x $3,433 = $11,700), which means students can do 70% of the program for 53% of the total cost. While we expect the majority of students who continue past the MicroMasters to finish the program, the program still enables a gradual financial commitment beyond the four classes. While the above mechanism does not offer refunds after a course has been taken (which under normal conditions is not done in any case for any degree at UCSD), the non-linear cost structure of the courses provides a very gradual financial commitment, which we believe compares favorably to the alternative of a linear cost structure with refunds. Put differently, if courses were offered at a linear cost scale, each course would cost $2,200, and the first four courses would cost $8,800. We instead offer those first four courses for $1,400, for all students, not just the ones who decide to leave after the MicroMasters. Within a quarter, we will use the following refund schedule. We follow a refund schedule similar to the Georgia Tech Online Masters in Computer Science (a top online program in computing from a top 10 school). Along with our schedule below, we also display the refund that UCSD in-person degrees would give at an equivalent point in time. Our proposed refund schedule is more generous than the standard in-person degree schedule.

% Refund if drop during ... MDS UCSD in Person

Week 1 100% 90%-100%

Week 2 100% 50%

Week 3 70% 25%-50%

Week 4 60% 25%

Week 5 50% 0%-25%

Week 6 40% 0%

Week 7 0% 0%

Week 8 0% 0%

Week 9 0% 0%

40

Week 10 0% 0%

Finally, it is important to realize that the refund process (and all aspects of the program) will inevitably be an iterative process. We will survey students, and understand what issues are causing problems for our students, and will consider solutions. If there are particular pain points that affect the welfare of our students, we will investigate, and address these pain points.

7.2. Diversity, Equity and Inclusion We will require each academic unit participating in the MDS to set aside at least 25% of their profit from MDS for the first three years to spend on Diversity, Equity, and Inclusion (DEI) for in-person programs. Note that profit is total revenue minus all expenses needed to run the program successfully. Academic units can choose to spend this 25% in creative ways, for example: funding diversity fellowships; funding outreach efforts; funding lecture series on diversity; giving stipends to faculty for large diversity initiatives. After three years of operation, a re-assessment will be made regarding the proportion of money that should be used toward DEI efforts, based on data collected during the first three years. In addition to this minimum of 25% used for DEI efforts for in-person programs, the MDS will also put money into DEI efforts for the online program itself, in the form of scholarships and outreach efforts.

Beyond the direct allocation of money to diversity, equity and inclusion, the online format itself makes the MDS program more accessible to underserved populations. There are three ways in which the online format achieves this goal:

1. The online format allows students who are not able to attend physically to get a Master degree; this includes non-local working professionals who must work while studying to make ends meet, international students not able to obtain visas, parents and caregivers who cannot attend an on-campus program and students with non-traditional backgrounds

2. The MicroMasters allows students from non-traditional backgrounds, who might otherwise be rejected, to demonstrate that they can do well in the program at a low cost

3. The online scalable format allows us to offer a degree at a much more affordable price point, thus making the program more financially accessible

The Program Director and Teaching Excellence Coordinator will also develop guidelines for inclusive and accessible teaching. There are already existing guidelines in the CSE department for in-person teaching, and we plan to adapt these guidelines to the online setting. The Program Director and Teaching Excellence Coordinator will ensure that these guidelines are adopted in all classes.

Finally, the program will work with the UCSD Office of Student Disabilities to ensure that students with disabilities are supported in the program.

41

8. Governance MDS is an interdisciplinary program. It will start with instructors from several units, including Computer Science and Engineering, Electrical and Computer Engineering, the San Diego Supercomputer Center, the Scripps Institution of Oceanography, and the Music Department. Over time, instructors from all over the UCSD campus will be involved in the program. HDSI will be the primary administrative home for the program. We recognize that at launch of MDS, the CSE department is the largest contributor of instructors and teaching assistants. Therefore, to ensure success and smooth operation, CSE will work closely with HDSI to administer the program for at least the first three years of operation. After the first three years, the governance structure will be reassessed and updated as needed by UCSD leadership and UCSD Academic Senate. Profits from the program will be distributed to the various academic units, in proportion to the participation of each academic unit in the teaching and course development in the program.

The program will be governed by the following administrative structure:

● Program Director: The director will be a member of the academic senate who will ultimately be responsible for the delivery of the program, and its ongoing evolution. For the first three years, the program director will report to the director of HDSI and the chair of CSE. After three years, this governance structure will be reassessed and updated as needed.

● Teaching Excellence Coordinator: Because maintaining quality of education is such an important consideration, a Teaching Excellence Coordinator (TEC) will be appointed. The TEC will be a member of the academic senate whose main responsibility will be to ensure that the program’s educational quality is of the highest caliber. The responsibilities of the TEC are detailed in Section 1.7. The TEC will report to the program director.

● Academic Integrity Coordinator: Because we want to enforce the highest level of academic integrity, an Academic Integrity Coordinator (AIC) will be appointed. The AIC will be a UCSD professor whose main role will be to ensure that all courses follow the guidelines set forth in Section 2.3.6, including running plagiarism detection software, and reporting all allegations of academic integrity violations. If needed, the efforts of the AIC will be supplemented with staff support, for example, to organize/manage/follow-up on all allegations of academic integrity violation.

● Admissions Committee: The admissions committee will consist of several Senate faculty members appointed by the program director. The admissions committee will decide which students are admitted into the program.

● Support Staff: The program will hire staff to run the following aspects of the program: admissions, academic and career advising, financial management, outreach. These staff will report to the program director.

42

9. Changes in Senate regulations The degree of Master of Data Science (online) will be granted on the following conditions.

A. Admission: The candidate shall have received a Bachelor’s degree in a relevant field; have fulfilled the requirements for admission to the Graduate Division of the University of California, San Diego; and shall meet any additional requirements that may be specified by the Admissions Committee of the program.

B. Residency: The minimum residence requirement is three academic quarters. Academic residence is met by satisfactory completion of at least four units in a quarter.

C. Scholarship: The candidate must maintain a 3.0 grade point average in all course work undertaken as a graduate student at the University of California.

D. Program: The program will consist of 40 units, including a capstone course, which is a quarter long project that applies all the previous content into a large project.

43

San Diego Program:Online

line # Yr 0 Yr 1 Yr 2 Yr 3 Yr 4 Yr 5

1 Year-average Program Enrollment (FTE) 100 150 200 300 400

2 Year-average Program Headcount 100 150 200 300 400

3 Annual Fee Per Student (Fee detail is optional)

Program Fee $21020 per yr $21020 per yr $21020 per yr $21020 per yr $21020 per yr

Student Services Fee TBD TBD TBD TBD TBD

Campus-Based Fees (SU+FL+SP) TBD TBD TBD TBD TBD

Health Insurance

Other (Example: Transit Fees) $0 $0 $0 $0 $0

4 Total Fee Revenue Generated $2,102,000 $3,153,000 $4,204,000 $6,306,000 $8,408,000

5 Total Other Funds $0 $0 $0 $0 $0

6 TOTAL PROGRAM REVENUE $2,102,000 $3,153,000 $4,204,000 $6,306,000 $8,408,000

Faculty FTE Equivalent* 0.60 0.60 0.60 0.60 0.60

Student-Faculty Ratio 166.67 250.00 333.33 500.00 666.67

7 Faculty Salaries (Course Delivery) & Benefits $75,000 $75,000 $75,000 $75,000 $75,000

8 Total Instructional Support Salaries & Benefits $537,281 $378,739 $299,468 $378,739 $458,010

9 Total Staff Salaries & Benefits $200,000 $204,000 $208,080 $212,242 $216,486

10 General Assistance

11 S&E

12 Equipment/Computing $15,000 $22,500 $30,000 $45,000 $60,000

13 Travel

14 Campus-based fee-funded activities (if any)

15 Other (Stipends, Course Maintenance) $90,000 $91,800 $93,636 $95,509 $97,419

16 TOTAL DIRECT COSTS, SUBJECT TO IDC $917,281 $772,039 $706,184 $806,489 $906,915

17 Financial Aid $206,000 $309,000 $412,000 $618,000 $824,000

18 Other S&E (describe _________________________________________) $0 $0 $0 $0 $0

19 Other Equipment (describe ____________________________________) $0 $0 $0 $0 $0

20 Other (describe Ex:____________) $0 $0 $0 $0 $0

21 TOTAL DIRECT COSTS, EXEMPT FROM IDC $206,000 $309,000 $412,000 $618,000 $824,000

Increase -3.8% 3.4% 27.4% 21.5%

22 TOTAL SCHOOL DIRECT COSTS (line 16 + line 21) $1,123,281 $1,081,039 $1,118,184 $1,424,489 $1,730,915

ANNUAL COST PER FTE STUDENT

23 Program Direct Costs (line 22 / line 1) $11,233 $7,207 $5,591 $4,748 $4,327

24 Program IDC Proxy (under UCSD SAPD ) 11.00% 22.00% 33.00% 33.00% 33.00%

25 Program Dean Proxy (under UCSD SAPD ) 7.00% 7.00% 7.00% 7.00% 7.00%

26 Program Indirect Costs Proxy (Campus+Dean Share under UCSD SAPD) $3,988 $6,254 $8,520 $8,520 $8,520

27 TOTAL COST PER FTE STUDENT $15,221 $13,461 $14,111 $13,269 $12,848

28 $1,522,081 $2,019,183 $2,822,264 $3,980,609 $5,139,075

29 $579,919 $1,133,818 $1,381,736 $2,325,391 $3,268,925

30 $5,799 $7,559 $6,909 $7,751 $8,172

Development Costs (Program)

31 New course and content 260,000$

Development Costs (Campus)

32 New course development/production 960,000$

Cost Analysis for Self-Supporting Program Fee Proposals Department-based SSGPDP

SURPLUS (DEFICIT) PER HEADCOUNT STUDENT

COSTS

REVENUE

Online Master of Data Science

FTE ENROLLMENT

Location:Campus:

A. Program Direct Costs, Subject to IDC (provided by the School)

SURPLUS (DEFICIT) (line 6 minus line 27)

TOTAL PROGRAM COST (line 1 x line 26)

B. Program Direct Costs, Exempt from IDC (provided by the School)

C. Total Direct Costs

Appendix A

UNIVERSITY OF CALIFORNIA, SAN DIEGO UCSD

BERKELEY DAVIS IRVINE LOS ANGELES MERCED RIVERSIDE SAN DIEGO SAN FRANCISCO

SANTA BARBARA SANTA CRUZ

OFFICE OF THE EXECUTIVE VICE CHANCELLOR

ACADEMIC AFFAIRS

9500 GILMAN DRIVE

LA JOLLA, CALIFORNIA 92093-0001

PHONE (858) 534-3130

FAX: (858) 534-5355

October 26, 2018

Sorin Lerner, Chair Graduate Council SUBJECT: Professional Master of Data Science This letter confirms my support for the proposed fully online Professional Master of Data Science as part of a self-supporting professional graduate degree program in Data Science administered by the Halicioğlu Data Science Institute. The Office of the Executive Vice Chancellor will provide the resource commitments via the use of the infrastructure in the Digital Learning Hub and Educational Technology Services; these two units are funded to enhance and support digital learning projects at UC San Diego and are equipped to support the needs of the proposed Master of Data Science. With best regards,

Elizabeth H. Simmons Executive Vice Chancellor, Academic Affairs Cc: Robert E. Continetti, Sr. AVC Academic Affairs

UNIVERSITY OF CALIFORNIA, SAN DIEGO

BERKELEY • DAVIS • IRVINE • LOS ANGELES • RIVERSIDE • SAN DIEGO • SAN FRANCISCO SANTA BARBARA • SANTA CRUZ

DEPARTMENT OF ELECTRICAL & COMPUTER ENGINEERING, 0407 LA JOLLA, CALIFORNIA 92093-0407

Professor Truong Nguyen Phone: (858) 822-5554 E-Mail: tqn001@eng.ucsd.edu September 17, 2018 Dean Tullsen, Ph.D. Rajesh Gupta, Ph.D. Professor and Chair Founding Institute Director Computer Science and Engineering Dept. Halicioglu Data Science Institute University of California, San Diego University of California, San Diego 9500 Gilman Dr. 9500 Gilman Dr. La Jolla, CA 92093 La Jolla, CA 92093 Dear Dean and Rajesh, The Electrical and Computer Engineering Department is fully supportive of the proposed online Master in Data Science. It is very timely as there is a strong demand for data scientists from the industry. By leveraging the existing UCSD Data Science MicroMasters, the proposed online program reduces the number of new courses it has to create, as well as provides a unique pathway for non-traditional students to participate. The ECE Dept. graduate program has a specialization in Machine Learning and Data Science. We hope that our students are able to take your courses, for elective credits. We also welcome joint course development in the future. Good luck with this exciting online Master program. Best regards,

Truong Nguyen, Ph.D. Prof. & Chair

9/27/2018 UCSD Jacobs School of Engineering Mail - Online Masters in Data Science proposal -- requesting a letter

https://mail.google.com/mail/u/0?ik=068efd933b&view=pt&search=all&permmsgid=msg-f%3A1612712783654802370&simpl=msg-f%3A16127127836… 1/1

Sorin Lerner <lerner@eng.ucsd.edu>

Online Masters in Data Science proposal -- requesting a letter

Lei Ni <lni.math.ucsd@gmail.com> Wed, Sep 26, 2018 at 3:57 PMTo: Dean Tullsen <tullsen@cs.ucsd.edu>Cc: Peter Ebenfelt <pebenfelt@ucsd.edu>, Sorin Lerner <lerner@cs.ucsd.edu>

Dear Dean, I have heard back from our faculty. It seems to us that the program is complementary to other master programs on campus, and the online master programhas some potentials. Given the above we support this program. Sincerely, Lei [Quoted text hidden]

MICHAEL L NORMAN, DIRECTOR 9500 GILMAN DRIVE, MC0505 SAN DIEGO SUPERCOMPUTER CENTER LA JOLLA, CALIFORNIA 92093-0505 TELEPHONE: (858) 822-5461 DISTINGUISHED PROFESSOR, FAX: (858) 534-5056 DEPARTMENT OF PHYSICS

Date: September 27, 2018

To: Dean Tullsen

Chair, Computer Science and Engineering Department Rajesh Gupta Director, Halicioglu Data Science Institute

From: Michael Norman

Director, San Diego Supercomputer Center

Re: Proposal to create an online Masters in Data Science program Dear Dean and Rajesh, I write to express my strong support for the creation of an online Masters in Data Science program at UCSD. I have reviewed the proposal, and find it to be excellent in its conception and detailed in its design. SDSC has a strong interest in participating in this program on a continuing basis through course development and instruction. I am happy to see the participation of my Chief Data Science Officer, Dr. Ilkay Altintas, in the program. This proposal continues and significantly expands on the joint partnership established several years ago between SDSC and CSE on the creation and delivery of the MAS program in Data Science and Engineering. There are several compelling reasons to create an online Masters in Data Science program at UCSD: • Data science is an important field of study which UCSD is having an increasing presence in; • There is strong demand from industry to hire data scientists; • The online nature and the price point of the program will allow it to reach populations not

current served by UCSD, including: non-local working professionals who must work while studying to make ends meet, international students not able to obtain visas, parents and caregivers who cannot attend an on-campus program and students with non-traditional backgrounds.

• The UCSD Data Science MicroMasters (which already exists) is a unique pathway into the program: it allows students from non-traditional backgrounds, who might otherwise be rejected, to demonstrate that they can do well in the program;

• This program leverages long term investments that UCSD has made in Data Science and online education.

Even though the new online MS program may eventually supercede or replace the MAS DSE program which provides some revenue to SDSC, I will view it as positive evolution as it will allow SDSC to reach a broader audience, with a potentially greater financial upside. I congratulate you both on the creation of this initiative, and wish you success in getting it approved. Please let me know if I can be of any further assistance. Sincerely,

Michael L. Norman, Director San Diego Supercomputer Center

DR. KAREN FLAMMER DIRECTOR, DIGITAL LEARNING TEACHING + LEARNING COMMONS 9500 GILMAN DRIVE #0175-W LA JOLLA, CALIFORNIA 92093-0175 December 1, 2018 Subject: HDSI Proposed Master of Data Science To: UC San Diego Academic Sentate I fully support the proposed online Master of Data Science which builds upon the existing MicroMasters program in Data Science. We are working closely with the HDSI/CSE faculty to ensure the courses within the MicroMasters are of the highest quality and will continue to do so as we support the design and approval of the remaining courses in the degree program. Furthermore, in order to provide a measure of consistent quality for online courses, UC San Diego is a subscribing member to Quality Matters TM , a world recognized quality assurance organization for online learning. The Digitial Learning team utilizes the Quality Matters Rubric to guide course design and facilitate a final course review. Sincerely,

Karen Flammer, Ph.D. kflammer@ucsd.edu

Alin Deutsch

Department of Computer Science and EngineeringUniversity of California San Diego9500 Gilman DriveLa Jolla, CA 92093-0404

phone: (858) 822-2276fax: (858) 534-7029

deutsch@cs.ucsd.edu

http://db.ucsd.edu/People/alin

EmploymentUniversity of California, San Diego, CA (July 2002 - present)Full Professor (since July 2012)Associate Professor (July 2008-June 2012)Assistant Professor (July 2002-June 2008)

Technical University of Darmstadt, Darmstadt, Germany (1994 - 1996)Research Assistant

EducationPh.D., Computer Science, University of Pennsylvania, 2002M.S. (“Hauptdiplom”), Computer Science, Technical University of Darmstadt, Germany, 1995B.S. (“Diploma”), Computer Engineering, University “POLITEHNICA” Bucharest, Romania, 1994

Awards/Honors

2018 ACM PODS Alberto O. Mendelzon Test of Time Award2018 Jean D’Alembert Fellowship from Universite Paris Saclay2006 Alfred P. Sloan Fellowship2006 Best SIGMOD Paper Award honorable mention (top-3 SIGMOD’06 papers)2004 NSF CAREER Award

Patents“A Systematic Approach to Query Optimization”, with Lucian Popa, Arnaud Sahuguet and Val Tan-nen. US Patent No. 6567802, awarded 5/20/2003.

Professional Service

• Demo Chair (with Nesime Tatbul) of International Conference on Very Large Databases (VLDB),2019

• Tutorial Chair (with Bertram Ludaescher) of International Conference on Data Engineering (ICDE),2017

• PC Chair of International Conference on Database Theory (ICDT), 2012

• Co-Chair (with Gavin Bierman) of ACM SIGPLAN Workshop on Programming Language Technolo-gies for XML (PLAN-X), 2009

• Co-Chair (with Wenfei Fan) of International Workshop on the Web and Databases (WebDB), 2006

• Program Committee member for

Symposium on Principles of Database Systems (PODS), 2019.ACM International Conference on Management of Data (SIGMOD), 2018.ACM International Conference on Management of Data (SIGMOD), 2017.International Conference on Business Process Management (BPM), 2017.Symposium on Principles of Database Systems (PODS), 2016.International Conference on Business Process Management (BPM), 2016.

1

Symposium on Principles of Database Systems (PODS), 2013.ACM International Conference on Management of Data (SIGMOD), 2013.Symposium on Principles of Database Systems (PODS), 2012.International Conference on Database Theory (ICDT), 2012.Symposium on Principles of Database Systems (PODS), 2011.International Workshop on the Web and Databases (WebDB), 2011.International Conference on Database Theory (ICDT), 2011.International Conference on Business Process Management (BPM), 2011.Alberto Mendelzon International Workshop on Foundations of Data Management, 2011.International Conference on Business Process Management (BPM), 2010.Asian Conference on Intelligent Information and Database Systems (ASIIDS), 2010.International Conference on Database Theory (ICDT), 2010.International Conference on Data Engineering (ICDE), 2010.VLDB PhD Workshop, 2009.International Conference on Very Large Databases (VLDB)– research track, 2009.International Conference on Very Large Databases (VLDB)– experiments and analysis track, 2009.International Workshop on the Web and Databases (WebDB), 2009.International Conference on the World Wide Web (WWW), 2009.International Conference on Data Engineering (ICDE), 2009.International Conference on Information and Knowledge Management (CIKM), 2009.International Conference on Advances in Databases and Information Systems (ADBIS), 2009.International Conference on Very Large Databases (VLDB), 2008.Symposium on Principles of Database Systems (PODS), 2008.International Workshop on Programming Languages for XML (PLAN-X), 2008.International Conference on Information and Knowledge Management (CIKM), 2008.International ICST Conference on Scalable Information Systems (InfoScale), 2008.International Conference on Data Engineering (ICDE), 2008.International Conference on Very Large Databases (VLDB), 2007.International Conference on Database Theory (ICDT), 2007.International Workshop on the Web and Databases (WebDB), 2007.International Workshop on Database and Programming Languages (DBPL), 2007.International Conference on Information and Knowledge Management (CIKM), 2007.ICDT Workshop on Emerging Research Opportunities in Web Data Management (EROW), 2007International Semantic Web Conference (ISWC), 2007.International Conference on Web-Age Information Management (WAIM), 2007.ACM International Conference on Management of Data (SIGMOD), 2006.Symposium on Principles of Database Systems (PODS), 2006.International Conference on Very Large Databases (VLDB), 2006.International Conference on Data Engineering (ICDE), 2006.International XML Database Symposium (XSym), 2006.International Conference on Information and Knowledge Management (CIKM), 2006.International Workshop on Web Information and Data Management (WIDM), 2006.Brazilian Symposium on Databases (SBBD), 2006.International Workshop on Database and Programming Languages (DBPL), 2005.International Conference on Web Services (ICWS), 2005.International Conference on Web-Age Information Management (WAIM), 2005.International Workshop on Data Integration in the Life Sciences (DILS), 2005.Asian Computing Science Conference on Data Management on the Web (ASIAN), 2005.VLDB PhD Workshop, 2005.International Conference on Very Large Databases (VLDB), 2004.International Conference on Data Engineering (ICDE), 2004.International Workshop on XQuery Implementation (XIME-P), 2004.International XML Database Symposium (XSym), 2004.International Workshop on the Web and Databases (WebDB), 2003.

2

International Conference on Information and Knowledge Management (CIKM), 2003.International Conference on Human.Society@Internet, 2003.

• Proceedings editor, Symposium on Principles of Database Systems (PODS), 2004.

Journal Publications1. “Automatic Verification of Database-Centric Systems”, Alin Deutsch, Richard Hull, Yuliang Li and

Victor Vianu. In SIGLOG News 5(2): 37-56, 2018.

2. “Automatic Verification of Database-Centric Systems”, Alin Deutsch, Richard Hull and Victor Vianu.In SIGMOD Record 43(3): 5-17, 2014.

3. “Abstractions for genomics”, Vineet Bafna, Alin Deutsch, Andrew Heiberg, Christos Kozanitis, LucilaOhno-Machado and George Varghese. In Communications of the ACM 56(1): 83-93, 2013.

4. “Automating the Database Schema Evolution Process”, Carlo Curino, Hyun Jin Moon, Alin Deutschand Carlo Zaniolo. In VLDB Journal 22(1): 73-98, 2013.

5. “Artifact systems with data dependencies and arithmetic”, Elio Damaggio, Alin Deutsch and VictorVianu. In ACM Transactions on Database Systems 37(3): 22, 2012.

6. “ASTERIX: towards a scalable, semistructured data platform for evolving-world models”, AlexanderBehm, Vinayak R. Borkar, Michael J. Carey, Raman Grover, Chen Li, Nicola Onose, Rares Vernica,Alin Deutsch, Yannis Papakonstantinou and Vassilis J. Tsotras. In Distributed and Parallel Databases29(3): 185-216, 2011.

7. “Querying XML data sources that export very large sets of views”, Bogdan Cautis, Alin Deutsch,Nicola Onose and Vasilis Vassalos. In ACM Transactions on Database Systems (TODS) 36(1), 2011.

8. “Querying Data Sources that Export Infinite Sets of Views”, Bogdan Cautis, Alin Deutsch and NicolaOnose. In Theory of Computing Systems (TOCS). 49(2): 367-428, 2011.

9. “WAVE: Automatic Verification of Data-Driven Web Services”, Alin Deutsch and Victor Vianu. InIEEE Data Engineering Bulletin 31(3), 2008.

10. “Interactive Query Formulation over Web Service-Accessed Sources: The CLIDE System”, MichalisPetropoulos, Alin Deutsch and Yannis Papakonstantinou. In ACM Transactions on Database Systems(TODS) 32(4), 2007.

11. “Rewriting Queries Using Views with Access Patterns Under Integrity Constraints”, Alin Deutsch,Bertram Ludaescher and Alan Nash. In Theoretical Computer Science (TCS), 371(3):200-226, 2007.

12. “Specification and Verification of Data-Driven Web applications”, Alin Deutsch, Liying Sui and VictorVianu. In Computer and Systems Sciences (JCSS), 73(3):442-474, 2007.

13. “Query Reformulation with Constraints”, Alin Deutsch, Lucian Popa and Val Tannen. In SIGMODRecord (Database Principles Column) 35(1):65-73, 2006.

14. “XML Queries and Constraints, Containment and Reformulation”, Alin Deutsch and Val Tannen. InTheoretical Computer Science (TCS) 336(1): 57-87 (2005).

15. “A Query Language for XML“, Alin Deutsch, Mary F. Fernandez, Daniela Florescu, Alon Y. Levyand Dan Suciu. In Computer Networks 31(11-16), 1999.

16. “Querying XML Data”, Alin Deutsch, Mary F. Fernandez, Daniela Florescu, Alon Y. Levy, DavidMaier and Dan Suciu. In IEEE Data Engineering Bulletin 22(3), 1999.

Refereed Conference Publications

17. ”Graph Data Models, Query Languages and Programming Paradigms”, Alin Deutsch and YannisPapakonstantinou. In International Conference on Very Large Databases (VLDB), Rio de Janiero,Brazil, 2018.

3

18. ”VERIFAS: A Practical Verifier for Artifact Systems”, Yuliang Li, Alin Deutsch and Victor Vianu.In International Conference on Very Large Databases (VLDB), Rio de Janiero, Brazil, 2018.

19. “Datalography: Scaling datalog graph analytics on graph processing systems”, Walaa Eldin Moustafa,Vicky Papavasileiou, Ken Yocum and Alin Deutsch. In BigData Conference, Washington DC, 2016.

20. “Flexible hybrid stores: Constraint-based rewriting to the rescue”, Francesca Bugiotti, Damian Bursz-tyn, Alin Deutsch, Ioana Manolescu and Stamatis Zampetakis. In International Conference on DataEngineering (ICDE), Helsinki, Finland, 2016.

21. “Towards a Shared Ledger Business Collaboration Language Based on Data-Aware Processes”, RichardHull, Vishal S. Batra, Yi-Min Chen, Alin Deutsch, Fenno F. Terry Heath III and Victor Vianu. InInternational Conference on Service-Oriented Computing (ICSOC), Banff, Canada, 2016.

22. “Verification of Hierarchical Artifact Systems”, Alin Deutsch, Yuliang Li and Victor Vianu. In Sym-posium on Principles of Database Systems (PODS), San Francisco, CA, 2016.

23. “Invisible Glue: Scalable Self-Tunning Multi-Stores”, Francesca Bugiotti, Damian Bursztyn, AlinDeutsch, Ioana Ileana and Ioana Manolescu. In Conference on Innovative Data Systems Research(CIDR), Asilomar, CA, 2015.

24. “Toward Scalable Hybrid Stores”, Francesca Bugiotti, Damian Bursztyn, Alin Deutsch, Ioana Ileana,and Ioana Manolescu. In Symposium on Advanced Database Systems (SEBD), Gaeta, Italy, 2015.

25. “State-Boundedness in Data-Aware Dynamic Systems”, Babak Bagheri Hariri, Diego Calvanese,Marco Montali and Alin Deutsch. In International Conference on the Principles of Knowledge Rep-resentation and Reasoning (KR), Vienna, Austria, 2014.

26. “Complete yet practical search for minimal query reformulations under constraints”, Ioana Ileana,Bogdan Cautis, Alin Deutsch, and Yannis Katsis. In ACM International Conference on Managementof Data (SIGMOD), Snowbird, UT, 2014.

27. “Verification of relational data-centric dynamic systems with external services”, Babak Bagheri Hariri,Diego Calvanese, Giuseppe De Giacomo, Alin Deutsch and MarAlin Deutsch, Yuliang Li, VictorVianu: co Montali. In Symposium on Principles of Database Systems (PODS), NYC, NY, 2013.

28. “Score-consistent algebraic optimization of full-text search queries with GRAFT”, Nathan Bales, AlinDeutsch, and Vasilis Vassalos. In ACM International Conference on Management of Data (SIGMOD),Athens, Greece, 2011.

29. “Querying contract databases based on temporal behavior”, Elio Damaggio, Alin Deutsch, and DayouZhou. In ACM International Conference on Management of Data (SIGMOD), Athens, Greece, 2011.

30. “Automatic Verification of Data-Centric Business Processes”, Elio Damaggio, Alin Deutsch, RichardHull and Victor Vianu. In International Conference on Business Process Management (BPM),Clermont-Ferrand, France, 2011.

31. “Artifact systems with data dependencies and arithmetic”, Elio Damaggio, Alin Deutsch, and VictorVianu. In International Conference on Database Theory (ICDT), Uppsala, Sweden, 2011.

32. “Update Rewriting and Integrity Constraint Maintenance in a Schema Evolution Support System:PRISM++”, Carlo Curino, Hyun Jin Moon, Alin Deutsch and Carlo Zaniolo. In International Con-ference on Very Large Databases (VLDB), Seattle, WA, 2011.

33. “Load-balanced query dissemination in privacy-aware online communities”, Emiran Curtmola, AlinDeutsch, K. K. Ramakrishnan, and Divesh Srivastava. In ACM International Conference on Man-agement of Data (SIGMOD), Indianapolis, IN, 2010.

4

34. “Policy-aware sender anonymity in location based services”, Alin Deutsch, Richard Hull, AvinashVyas, and Kevin Keliang Zhao. In International Conference on Data Engineering (ICDE), LongBeach, CA, 2010.

35. “Inconsistency resolution in online databases”, Yannis Katsis, Alin Deutsch, Yannis Papakonstantinouand Vasilis Vassalos. In demo track of International Conference on Data Engineering (ICDE), LongBeach, CA, 2010.

36. “Efficient Rewriting of XPath Queries Using Query Set Specifications”, Bogdan Cautis, Alin Deutsch,Nicola Onose and Vasilis Vassalos. In International Conference on Very Large Databases (VLDB),Lyon, France, 2009.

37. “FORWARD: Design Specification Techniques for Do-It-Yourself Application Platforms”, GauravBhatia, Yupeng Fu, Keith Kowalczykowski, Kian Win Ong, Kevin Keliang Zhao, Alin Deutsch, andYannis Papakonstantinou. In International Workshop on the Web and Databases (WebDB), Provi-dence, RI, 2009.

38. “Automatic verification of data-centric business processes”, Alin Deutsch, Richard Hull, Fabio Patrizi,and Victor Vianu. In International Conference on Database Theory (ICDT), St. Petersburg, Russia,2009.

39. “Querying data sources that export infinite sets of views”, Bogdan Cautis, Alin Deutsch and NicolaOnose. In International Conference on Database Theory (ICDT), St. Petersburg, Russia, 2009.

40. “Do-It-Yourself custom forms-driven workflow applications, Keith Kowalczykowski, Kian Win Ong,Kevin Keliang Zhao, Alin Deutsch, Yannis Papakonstantinou and Michalis Petropoulos. In Conferenceon Innovative Data Systems Research (CIDR), Asilomar, California, 2009.

41. “The chase revisited”, Alin Deutsch, Alan Nash and Jeffrey B. Remmel. In Symposium on Principlesof Database Systems (PODS), Vancouver, Canada, 2008.

42. “XPath Rewriting Using Multiple Views: Achieving Completeness and Efficiency”, Bogdan Cautis,Alin Deutsch and Nicola Onose. In International Workshop on the Web and Databases (WebDB),Vancouver, Canada, 2008.

43. “XTreeNet: democratic community search”, Emiran Curtmola, Alin Deutsch, Dionysios Logothetis,K. K. Ramakrishnan, Divesh Srivastava and Ken Yocum. In demo track of International Conferenceon Very Large Databases (VLDB), Auckland, New Zealand, 2008.

44. “RIDE: a tool for interactive source registration in community-oriented information integration”,Yannis Katsis, Alin Deutsch, Yannis Papakonstantinou and Kevin Keliang Zhao. In demo track ofInternational Conference on Very Large Databases (VLDB), Auckland, New Zealand, 2008.

45. “Managing and querying transaction-time databases under schema evolution.”, Hyun J. Moon, CarloCurino, Alin Deutsch, Chien-Yi Hou and Carlo Zaniolo. In International Conference on Very LargeDatabases (VLDB), Auckland, New Zealand, 2008.

46. “Interactive source registration in community-oriented information integration”, Yannis Katsis, AlinDeutsch and Yannis Papakonstantinou. In International Conference on Very Large Databases (VLDB),Auckland, New Zealand, 2008.

47. “CLIDE: Interactive Query Formulation for Service-Oriented Architectures”, Michalis Petropoulos,Alin Deutsch and Yannis Papakonstantinou. In demo track of ACM International Conference onManagement of Data (SIGMOD), Beijing, China, 2007.

48. “Privacy in GLAV Information Integration”, Alan Nash and Alin Deutsch. In International Confer-ence on Database Theory (ICDT), Barcelona, Spain, 2007.

5

49. “Interactive Query Formulation Over Web Service-Accessed Sources”, Michalis Petropoulos, AlinDeutsch and Yannis Papakonstantinou. In ACM International Conference on Management of Data(SIGMOD), Chicago, Illinois, 2006. Top-3 Best Paper Award.

50. “Flexible and Efficient XML Search with Complex Full-Text Predicates”, Sihem Amer-Yahia, EmiranCurtmola and Alin Deutsch. In ACM International Conference on Management of Data (SIGMOD),Chicago, Illinois, 2006.

51. “Rewriting Nested XML Queries Using Nested Views”, Nicola Onose, Alin Deutsch, Yannis Pa-pakonstantinou and Emiran Curtmola. In ACM International Conference on Management of Data(SIGMOD), Chicago, Illinois, 2006.

52. “A System for Specification and Verification of Data-driven Web Applications”, Alin Deutsch, LiyingSui, Victor Vianu and Dayou Zhou. In demo track of ACM International Conference on Managementof Data (SIGMOD), Chicago, Illinois, 2006.

53. “Verification of Communicating Data-Driven Web Services”, Alin Deutsch, Liying Sui, Victor Vianuand Dayou Zhou. In Symposium on Principles of Database Systems (PODS), Chicago, Illinois, 2006.

54. “A Verifier for Interactive, Data-Driven Web Applications”, Alin Deutsch, Monica Marcus, LiyingSui, Victor Vianu and Dayou Zhou. In ACM International Conference on Management of Data(SIGMOD), Baltimore, Maryland, 2005.

55. “Determining Source Contribution in Information Integration Systems”, Alin Deutsch, Yannis Katsisand Yannis Papakonstantinou. In Symposium on Principles of Database Systems (PODS), Baltimore,Maryland, 2005.

56. “Privacy in Database Publishing”, Alin Deutsch and Yannis Papakonstantinou. In InternationalConference on Database Theory (ICDT), Edinburgh, UK, 2005.

57. “Rewriting Queries Using Views with Access Patterns Under Integrity Constraints”, Alin Deutsch,Bertram Ludaescher and Alan Nash. In International Conference on Database Theory (ICDT), Ed-inburgh, UK, 2005.

58. “The Role of Visual Tools in a Web Application Design and Verification Framework: A Visual Notationfor LTL Formulae”, Marco Brambilla, Alin Deutsch, Liying Sui and Victor Vianu. In InternationalConference on Web Engineering (ICWE), Sydney, Australia, 2005.

59. “Building an XQuery interpreter in a compiler construction course”, Sara Miner More, Tim Pevzner,Alin Deutsch, Scott B. Baden and Paul Kube. In Technical Symposium on Computer Science Educa-tion (SIGCSE), St. Louis, Missouri, 2005.

60. “Integrating XML Data Sources using RDF/S Schemas: The ICS-FORTH Semantic Web IntegrationMiddleware (SWIM)”, Ioanna Koffina, Giorgos Serfiotis, Vassilis Christophides, Val Tannen and AlinDeutsch. In Proceedings of Dagstuhl Seminar on Semantic Interoperability and Integration, Dagstuhl,Germany, 2005.

61. “The NEXT Logical Framework for XQuery”, Alin Deutsch, Yannis Papakonstantinou and Yu Xu.In International Conference on Very Large Databases (VLDB), Toronto, Canada, 2004.

62. “Specification and Verification of Data-driven Web Services”, Alin Deutsch, Liying Sui and VictorVianu. In Symposium on Principles of Database Systems (PODS), Paris, France, 2004.

63. “Minimization and Group-By Detection for Nested XQueries”, Alin Deutsch, Yannis Papakonstanti-nou and Yu Xu. Poster in International Conference on Data Engineering (ICDE), Boston, MA,2004.

64. “MARS: A System for Publishing XML from Mixed and Redundant Storage”, Alin Deutsch and ValTannen. In International Conference on Very Large Databases (VLDB), Berlin, Germany, 2003.

6

65. “Reformulation of XML Queries and Constraints”, Alin Deutsch and Val Tannen. In InternationalConference on Database Theory (ICDT), Siena, Italy, 2003.

66. “The Query Set Specification Language (QSSL)”, Michalis Petropoulos, Alin Deutsch and Yannis Pa-pakonstantinou. In International Workshop on Web and Databases (WebDB), San Diego, California,2003.

67. “Containment and Integrity Constraints for XPath Fragments”, Alin Deutsch and Val Tannen. InInternational Workshop on Knowledge Representation meets Databases (KRDB), Rome, Italy, 2001.

68. “Optimization Properties for Classes of Conjunctive Regular Path Queries,”, Alin Deutsch and ValTannen. In International Workshop on Database Programming Languages (DBPL), Frascati, Italy,2001.

69. “A Chase Too Far?”, Lucian Popa, Alin Deutsch, Arnaud Sahuguet and Val Tannen. In ACM Inter-national Conference on Management of Data (SIGMOD), Dallas, Texas, 2000.

70. “Physical Data Independence, Constraints and Optimization with Universal Plans”, Alin Deutsch,Lucian Popa and Val Tannen. In International Conference on Very Large Databases (VLDB), Edin-burgh, Scotland, 1999.

71. “Storing Semistructured Data with STORED”, Alin Deutsch, Mary Fernandez and Dan Suciu. InACM International Conference on Management of Data (SIGMOD), Philadelphia, Pennsylvania,1999.

72. “A Query Language for XML“, Alin Deutsch, Mary F. Fernandez, Daniela Florescu, Alon Y. Levyand Dan Suciu. In International World-Wide-Web Conference (WWW), 1999.

73. “A Deterministic Model for Semistructured Data”, Peter Buneman, Alin Deutsch and Wang-ChiewTan. In Workshop on Query Processing for Semistructured Data and Non-Standard Data Formats (inconjunction with ICDT’99), Jerusalem, Israel, 1998.

74. “Beyond XML Query Languages”, Peter Buneman, Alin Deutsch, Wenfei Fan, Hartmut Liefke,Arnaud Sahuguet and Wang-Chiew Tan. In Workshop on Query Languages (QL), Boston, Mas-sachusetts, 1998.

75. “Design, Implementation and Management of Rules in an Active Database System”, Juergen Zimmer-mann, Holger Brandig, Alejandro P. Buchmann, Alin Deutsch and Andreas Geppert. In InternationalConference on Database and Expert Systems Applications (DEXA), Zurich, Switzerland, 1996.

76. “The REACH Active OODBMS”, Alejandro P. Buchmann, Alin Deutsch, Juergen Zimmermann andM. Higa. In demo track of ACM International Conference on Management of Data (SIGMOD), SanJose, California, 1995.

Book Chapters

77. ”Provenance-Directed Chase & Backchase”, Alin Deutsch and Richard Hull. In In Search of Elegancein the Theory and Practice of Computation, Springer Verlag, 2013.

78. “FOL Modeling of Integrity Constraints (Dependencies)”, Alin Deutsch. In Ling Liu and M. TamerOszu (Eds), Encyclopedia of Database Systems, pages 1155-1161, Springer Verlag, 2009.

79. “Chase”, Alin Deutsch and Alan Nash. In Ling Liu and M. Tamer Oszu (Eds), Encyclopedia ofDatabase Systems, pages 323-327, Springer Verlag, 2009.

80. “Privacy in Database Publishing: a Bayesian Perspective”, Alin Deutsch. In Michael Gertz and SushilJajodia (Eds), Handbook of Database Security: Applications and Trends. Springer Verlag, 2007.

Other Publications

7

81. “XML-QL: A Query Language for XML”, Alin Deutsch, Mary Fernandez, Daniela Florescu, AlonLevy and Dan Suciu. World-Wide-Web Consortium (W3C) Note xml-ql-19980819, August 1998.

Software in the Public Domain

The XML-QL engine (downloadable for several years from the AT&T Research Lab Web site). BetweenDec. 1998 and Dec. 2001, the software was downloaded by 2672 distinct users from 2219 distinctcompanies.

8

Ilkay Altintas San Diego Supercomputer Center Telephone: (858) 822-5453 9500 Gilman Drive Fax: (858) 822-3693 MC 0505 E-mail: altintas@sdsc.edu La Jolla, CA 92093-0505 Professional Preparation

Middle East Technical University, Ankara, Turkey

B.S. Computer Engineering 1999

Middle East Technical University, Ankara, Turkey

M.S. Computer Engineering 2001

University of Amsterdam, Amsterdam, Netherlands

Ph.D. Computational Science 2011

Appointments 2016-. Associate Research Scientist, San Diego Supercomputer Center, UCSD

2016-. 2015-. 2015-. 2014-.

Faculty Co-Director, Master of Advanced Studies in Data Science and Engineering, UCSD Chief Data Science Officer, San Diego Supercomputer Center (SDSC), UCSD Division Director, Cyberinfrastructure Research, Education and Development, SDSC, UCSD Founder and Director, Workflows for Data Science Center of Excellence, SDSC, UCSD

2012-. 2012-2016

Lecturer, Department of Computer Science and Engineering, UCSD Assistant Research Scientist, San Diego Supercomputer Center, UCSD

2008-2014 Deputy Coordinator for Research, San Diego Supercomputer Center, UCSD

2004-2014 Founder and Director, Scientific Workflow Automation Technologies Laboratory, SDSC, UCSD

2005-2007 Assistant Director, National Laboratory for Advanced Data Research (NLADR) - Data, SDSC, UCSD

2001-2004 Research Programmer (P/A III), SDSC, UCSD

1999-2001 Research Assistant, Middle East Technical University (Ankara, TURKEY)

Products (Out of 100+) 1. I. Altintas, J. Block, R. de Callafon, D. Crawl, C. Cowart, A. Gupta, M.Nguyen, H.W. Braun, J.

Schulze, M. Gollner, A. Trouve, L. Smarr: Towards an Integrated Cyberinfrastructure for Scalable Data-Driven Monitoring, Dynamic Prediction and Resilience of Wildfires. In Proceedings of the Workshop on Dynamic Data-Driven Application Systems (DDDAS) at the 15th International Conference on Computational Science (ICCS 2015), Procedia Computer Science, Volume 51, 2015, Pages 1633-1642, ISSN 1877-0509, doi:10.1016/j.procs.2015.05.296. (Best Paper Award)

2. Kepler Scientific Workflow System Releases 1.0, 2.0 through 2.4. (Downloaded by 100K+) 3. J. Wang, D. Crawl, I. Altintas, W. Li. Big Data Applications using Workflows for Data Parallel

Computing. Computing in Science & Eng., 16(4), pp. 11-22, July-Aug. 2014, IEEE.

4. J. Wang, P. Korambath, I. Altintas, J. Davis, D. Crawl. Workflow as a Service in the Cloud: Architecture and Scheduling Algorithms. In Proceedings of International Conference on Computational Science (ICCS 2014), pages 546-556. DOI: 10.1016/j.procs.2014.05.049

5. B. Ludaescher, I. Altintas, C. Berkley, D. Higgins, E. Jaeger-Frank, M. Jones, E. Lee, J. Tao, Y. Zhao, Scientific Workflow Management and the Kepler System, Concurrency and Computation: Practice & Experience, 18(10), pp. 1039-1065, 2006. (Cited by 1952 in January 2018.)

Other Selected Products 6. I. Altintas, M.K. Anand, T. Vuong, S. Bowers, B. Ludaescher, P.M.A. Sloot, “A Data Model for

Analyzing User Collaborations in Workflow-Driven eScience,” The International Journal of Computers and Their Applications (IJCA), 2011. Vol. 18, No. 3, p.160 – 180, Dec, 2011.

7. I. Altintas, A.W. Lin, J. Chen, C. Churas, M. Gujral, S. Sun, W. Li, R. Manansala, M. Sedova, J.S. Grethe, and M. Ellisman, “CAMERA 2.0: A Data-centric Metagenomics Community Infrastructure Driven by Scientific Workflows,” In Proceedings of the SWF 2010 at IEEE SERVICES '10, pp. 352-359, 2010. DOI=10.1109/SERVICES.2010.89

8. A. Goderis, C. Brooks, I. Altintas, E. Lee, and C. Goble, “Heterogeneous composition of models of computation,” FGCS, vol. 25, no. 5, pp. 552–560, 2009.

9. I.Altintas, O. Barney, E. Jaeger-Frank, Provenance Collection Support in the Kepler Scientific Workflow System, in Provenance and Annotation of Data, LNCS Volume 4145/2006, pages 118-132, 2006. (Cited by 314 in January 2018.)

10. I. Altintas, C. Berkley, E. Jaeger, M. Jones, B. Ludaescher, and S. Mock, “Kepler: An extensible system for design and execution of scientific workflows,” in Intl. Conference on Scientific and Statistical Database Management (SSDBM), Greece, 2004. (Cited by 980 in January 2018.)

Recent Synergistic Activities and Awards • Recent Honors: CENIC 2018 Innovations in Networking Award for Experimental Applications,

2018; ACM SIGHPC Emerging Woman Leader in Technical Computing Award, 2017; Peter Chen Big Data Young Researcher Award, 2017; IEEE TCSC Award for Excellence for Early Career Researchers, 2015; SDSC Pi Person of the Year, 2014; HPCwire Reader’s Choice Award “Best Application of Big Data in HPC”, 2014; HPCwire Reader’s Choice Award “Best Data-Intensive System”, 2014; HPCwire Editor’s Choice Awards, 2014; Best Workshop Paper Award, International Conference on Computational Science, 2015; Outstanding Teaching Service Award (nominated by UC San Diego students with disabilities in recognition of accommodations), 2015 and 2016.

• Some Memberships and Advisory Boards: ACM (2002-.), IEEE and WIC (2003-.), ORION: CI Committee (2004-2007), OMII-UK: Technical Advisory Group, AGU Earth and Space Science Informatics: Executive Committee, Data-Enabled Life Science Alliance CI Committee (Co-chair; 2012-2013), Kepler Collaboration Leadership

• Review Panelist: NSF (2007, 2010, 2011, 2013, 2014, 2015, 2016), DOE (2010, 2011, 2012, 2013, 2015), Genome Canada (2012), Skoltech Institute (2013)

• Recent Academic Service: Associate editor: Future Generation Computer Systems Journal, Elsevier (IF: 2.369); Issue Editor: Special Issue on Experimental Software Engineering in the Cloud (ESEiC), Science of Computer Programming Journal, Elsevier, 2012; Judge for the Elsevier Executable Paper Grand Challenge, 2011; Recent Program Committees: (2014) SWF (Chair), Big Data, e-Science , IPAW, ICCS; (2013) SWF (Chair), Big Data, e-Science , ICCS; (2012) e-Science, ICCS, IPAW, CloudFlow (also on Steering Committee); (2011) e-Science, ICCS, SWF (publicity chair); Journal Reviewer: VLDB (Journal Track), FGCS, CAGEO, SIMPAT, ACM SIGMOD Record, Concurrency and Computation: Practice and Experience.

Arun Kumar

3218 EBU3B (CSE building)9500 Gilman Drive, Mail Code 0404La Jolla, CA 92093

Email : arunkk@eng.ucsd.eduPhone: (+1) 614-602-9734Web: http://cseweb.ucsd.edu/~arunkk/

EMPLOYMENT University of California, San DiegoDepartment of Computer Science and EngineeringAssistant Professor 2016–Now

EDUCATION University of Wisconsin-MadisonPh.D. in Computer Sciences. 2011–2016Thesis Co-advisors: Jeffrey Naughton and Jignesh M. Patel

M.S. in Computer Sciences. 2009–2011Research Supervisor: Christopher Re

Indian Institute of Technology, MadrasB.Tech. in Computer Science and Engineering. 2005–2009

RESEARCHINTERESTS

Data management and its intersection with ML, especially devising data management-inspired abstractions, systems, frameworks, and algorithms to make the end-to-endprocess of building and deploying ML/AI algorithms for data analytics applicationseasier (improving the productivity of data scientists and developers) and faster (im-proving runtime performance and introducing accuracy trade-offs). My work spansthe gamut of building new data systems, algorithm design, empirical analysis, theo-retical analysis, and working with practitioners to help deploy my research.

Research Webpage: https://adalabucsd.github.io/

SELECTEDHONORS

Hellman Fellowship 2018Faculty of the Year from UCSD oSTEM Chapter 2018ACM SIGMOD Distinguished PC Member 2017Google Faculty Research Award 2017Invited Keynote at ACM SIGMOD DEEM Workshop 2017UW CS Graduate Student Research Award for best PhD research 2016Invited Paper at ACM Transactions on Database Systems 2016Anthony C. Klug NCR Fellowship in Database Systems 2015Best Paper Award at ACM SIGMOD 2014Invited Paper at the Communications of the ACM 2013National Talent Search Exam (NTSE) Scholarship by the Govt. of India 2003–08

CONFERENCEPUBLICATIONS

Model-based Pricing for Machine Learning in a Data MarketplaceL. Chen, P. Koutris, and A. KumarUnder submission

Materialization Trade-offs for Feature Transfer from Deep CNNs for Multimodal DataAnalyticsS. Nakandala and A. KumarUnder submission

Tuple-Oriented Compression for Large-scale Mini-Batch Gradient DescentF. Li, L. Chen, A. Kumar, J. Naughton, J. M. Patel, and X. Wu

Under submission

Hierarchical Machine Learning Inference for Networked Sensing ApplicationsA. Thomas, Y. Guo, Y. Kim, B. Aksanli, A. Kumar, and T. S. RosingUnder submission

A Comparative Evaluation of Systems for Scalable Linear Algebra-based AnalyticsA. Thomas and A. KumarVLDB 2018/2019 (To appear)

In-RDBMS Hardware Acceleration of Advanced AnalyticsD. Mahajan, J. K. Kim, J. Sacks, A. Ardalan, A. Kumar, and H. EsmaeilzadehVLDB 2018 (To appear)

Are Key-Foreign Key Joins Safe to Avoid when Learning High Capacity Classifiers?V. Shah, A. Kumar, and X. ZhuVLDB 2018 (To appear)

Towards Linear Algebra over Normalized DataL. Chen, A. Kumar, J. Naughton, and J. M. PatelVLDB 2017

Bolt-on Differential Privacy for Scalable Stochastic Gradient Descent-based AnalyticsX. Wu, F. Li, A. Kumar, K. Chaudhuri, S. Jha, and J. NaughtonACM SIGMOD 2017

Cerebro: A System to Manage Deep Learning for Relational Data AnalyticsA. KumarCIDR 2017 (Abstract)

To Join or Not to Join? Thinking Twice about Joins before Feature SelectionA. Kumar, J. Naughton, J. M. Patel, and X. ZhuACM SIGMOD 2016

Learning Generalized Linear Models Over Normalized DataA. Kumar, J. Naughton, and J. M. PatelACM SIGMOD 2015

Materialization Optimizations for Feature Selection WorkloadsC. Zhang, A. Kumar, and C. ReACM SIGMOD 2014 (Best Paper Award; Invited to ACM TODS 2016)

Brainwash: A Data System for Feature EngineeringM. Anderson, D. Antenucci, V. Bittorf, M. Burgess, M. Cafarella, A. Kumar, F. Niu,Y. Park, C. Re, and C. ZhangCIDR 2013 (Vision paper)

Probabilistic Management of OCR Data Using an RDBMSA. Kumar, and C. ReVLDB 2012

The MADlib Analytics Library: Or MAD Skills, the SQLJ. Hellerstein, C. Re, F. Schoppmann, D. Wang, E. Fratkin, A. Gorajek, K. Ng, C.Welton, X. Feng, K. Li, and A. KumarVLDB 2012 (Industrial track)

Towards a Unified Architecture for in-RDBMS AnalyticsX. Feng*, A. Kumar*, B. Recht, and C. Re (*alphabetical order of surnames)ACM SIGMOD 2012

Mobile Data Collection in WSNs Using Wireless Communication

A. Kumar and K. M. SivalingamIEEE/ACM COMSNETS 2010

JOURNALPUBLICATIONS

Materialization Optimizations for Feature Selection WorkloadsC. Zhang, A. Kumar, and C. ReACM TODS 2016 (Invited paper)

Model Selection Management Systems: The Next Frontier of Advanced AnalyticsA. Kumar, R. McCann, J. Naughton, and J. M. PatelACM SIGMOD Record Dec 2015 (Vision paper)

On Reducing Delay in Mobile Data Collection-Based WSNsA. Kumar, K. M. Sivalingam, and A. KumarSpringer Wireless Networks 2012

OTHER PEER-REVIEWEDPUBLICATIONS

Model-based Pricing: Do Not Pay for More than What You Learn!L. Chen, P. Koutris, and A. KumarACM SIGMOD 2017 DEEM Workshop

SpeakQL: Towards Speech-driven Multi-modal QueryingD. Chandarana, V. Shah, A. Kumar, and L. SaulACM SIGMOD 2017 HILDA Workshop

Demonstration of Santoku: Optimizing Machine Learning over Normalized DataA. Kumar, M. Jalal, B. Yan, J. Naughton, and J. M. PatelVLDB 2015 (Demo)

Hazy: Making it Easier to Build and Maintain Big-data AnalyticsA. Kumar, F. Niu, and C. ReACM Queue 2013 (Invited to the Communications of the ACM)

Distributed and Scalable PCA in the CloudA. Kumar, N. Karampatziakis, P. Mineiro, M. Weimer, and V. NarayananNIPS BigLearn Workshop 2013

Feature Selection in Enterprise Analytics: A Demonstration using an R-based DataAnalytics SystemP. Konda, A. Kumar, C. Re, and V. SashikanthVLDB 2013 (Demo)

Flexible Multimedia Content Retrieval Using InfoNamesA. Kumar, A. Anand, A. Balachandran, V. Sekar, A. Akella, S. SeshanACM SIGCOMM 2010 (Demo)

TECHNICALREPORTS ANDMANUSCRIPTS

Learning Over JoinsA. KumarUW-Madison CS PhD Dissertation, 2016

A Survey of the Existing Landscape of ML SystemsA. Kumar, R. McCann, J. Naughton, and J. M. PatelUW-Madison CS Technical Report TR1827, 2015

InfoNames: An Information-Based Naming Scheme for Multimedia ContentA. Kumar, A. Anand, A. Balachandran, V. Sekar, A. Akella, S. SeshanUW-Madison CS Technical Report TR 1677, 2010

TEACHING CSE 190A: Topics in Database System ImplementationInstructor. UCSD. Spring 2018

CSE 291A: Advanced Data Analytics and ML SystemsInstructor. UCSD. Winter 2018

CSE 290A: Seminar on Advanced Data ScienceOrganizer. UCSD. Fall 2018

CSE 190D: Topics in Database System ImplementationInstructor. UCSD. Spring 2017

CSE 290B: Seminar on Advanced Data ScienceOrganizer. UCSD. Spring 2017

CSE 291G: Topics in Advanced AnalyticsInstructor. UCSD. Winter 2017

CS 564: Database Management Systems: Design and ImplementationInstructor. UW-Madison. Fall 2015

CS 764: Topics in Database Management SystemsGuest Lecture (Instructor: Jeffrey Naughton). UW-Madison. Fall 2015

CS 764: Topics in Database Management SystemsGuest Lecture (Instructor: Christopher Re). UW-Madison. Spring 2013

ADVISING(CURRENT)

Lingjiao Chen, PhD at UW-Madison (Co-advisor: Paris Koutris). Fall 2015–Supun Nakandala, PhD at UCSD. Fall 2017–Vraj Shah, MS & PhD at UCSD. Fall 2016–Side Li, BS at UCSD. Fall 2017–

ADVISING(ALUMNI)

Anthony Thomas, MS at UCSD. Winter 2017–Spring 2018Mingyang Wang, MS at UCSD. Spring 2017

THESISCOMMITTEE

Julaiti Alafate, PhD at UCSD (Advisor: Yoav Freund). 2018Title TBD

Chunbin Lin, PhD at UCSD (Advisor: Yannis Papakonstantinou). 2017“Accelerating Query Processing on Compressed Data”

Nishant Agarwal, MS at UCSD (Advisor: Amarnath Gupta). 2017“A Real-Time Temporal Clustering Algorithm for Short Text, and its Applications”

Sumedha Kattar, MS at UCSD (Advisor: Ilkay Altintas). 2017“Finding the burnability index of a point on a map using the historical fire data”

RESEARCHEXAMCOMMITTEE

Rana Alotaibi, PhD at UCSD (Advisor: Alin Deutsch). Spring 2018Title TBD

Nikos Koulouris, PhD at UCSD (Advisor: Yannis Papakonstantinou). Spring 2018“Controlling for False Discoveries in Data Exploration Systems”

SERVICE Organization:Co-Chair, ACM SIGMOD 2018 Workshop on Data Management for End-to-End ML

(DEEM)Organizing Committee, ACM SIGKDD 2018 Workshop on Common Model Infras-tructure (CMI)Organizing Committee, Extremely Large Databases (XLDB) Conference 2018

Program Committee:ACM SIGMOD 2019, 2018, 2017VLDB 2019, 2018ACM SIGMOD 2017 Demonstrations and Student Research CompetitionACM SIGMOD 2017 Workshop on Data Management for End-to-End ML (DEEM)IEEE ICDE 2017USENIX 2016 Workshop on Hot Topics in Cloud Computing (HotCloud)ACM SIGMOD 2016 Undergraduate Research Poster Competition

Reviewer:ACM Transactions on Database Systems (TODS) 2017ACM Transactions on Database Systems (TODS) 2015IEEE Transactions on Knowledge and Data Engineering (TKDE) 2014

External Reviewer:VLDB 2017, ACM SIGMOD 2013, IEEE ICDE 2013IEEE INFOCOM 2010, IEEE GLOBECOM 2009, IEEE SECON 2009

Other Research-Related:Speaker at ACM SIGMOD 2018 New Researcher SymposiumInterviewee for ACM SIGMOD 2018 WebDB Workshop Article on “Data meets ML”Co-chair of “Best of ICDE 2017” Selection Committee for TKDE 2018Judge for ACM SIGMOD 2017 Student Research CompetitionPanelist at IEEE ICDE 2017 PhD SymposiumJudge for IEEE ICDE 2017 Demonstrations

Outreach/Contributions to Diversity:Winter–Spring 2018: Member, UCSD LGBTQIA+ Undergraduate Scholarships Com-mitteeFall 2017–: Member of UCSD CSE Diversity CommitteeNov 2017: Attended the annual conference of oSTEM representing CSE and UCSDNov 2017: Panelist for a Q & A event organized by oSTEM UCSD chapter for outLGBTQ+ students in STEMOct 2017: Co-proposed new CSE PhD scholarship for contributions to diversityApr 2017: Spoke about my coming out experience in graduate school as a panelist atthe IEEE ICDE 2017 PhD SymposiumApr 2017: Part of the faculty group on diversity issues during CSE external reviewFall 2016–: Listed on the UCSD LGBT Resource Center “Out List” of faculty mentorsfor LGBTQ+ students

TALKS Accelerating Model Selection in Advanced AnalyticsTeradata, San Diego (Invited) Nov 2017Opera Solutions Technical Conference, San Diego (Invited) Oct 2017University of Michigan, Ann Arbor (Invited) Sep 2017

Towards Linear Algebra over Normalized DataVLDB Aug 2017

Accelerating Advanced Analytics on Multi-table DataAmazon Machine Learning, Berlin (Invited) Aug 2017

Democratizing Advanced Analytics Beyond Just PlumbingACM SIGMOD DEEM Workshop (Invited Academic Keynote) May 2017

Democratizing Feature Engineering and Model Selection in Advanced AnalyticsOpera Solutions, San Diego (Invited) May 2017

Democratizing Distributed Advanced AnalyticsUCSD Center for Networked Systems Lecture Apr 2017

Cerebro: A System to Manage Deep Learning for Relational Data AnalyticsCIDR “Gong Show” Jan 2017

Accelerating Advanced AnalyticsGoogle, Mountain View (Invited) Dec 2016

The Data Strikes Back! Research Challenges in Advanced AnalyticsUCSD AI Seminar Oct 2016

Exploiting Database Dependencies to Accelerate Advanced AnalyticsUCSD Database Seminar Oct 2016

Model-based Pricing of Relational Data in the CloudUCSD Database Seminar Oct 2016

Accelerating Advanced Analytics (Invited) Jan-Mar 2016

New York UniversityMicrosoft Research, Redmond, WAUniversity of Illinois at Urbana-ChampaignCornell UniversityUniversity of California, San Diego (Video: https://goo.gl/raJFpu)University of ChicagoIBM Research Almaden, CA (under a different title)University of Maryland, College ParkLogicBlox, Atlanta, GAGeorgia Institute of TechnologyPurdue University (under a different title)

Machine Learning over Joins of Multiple TablesWisconsin Institutes of Discovery Seminar 2015

Learning Generalized Linear Models over Normalized DataACM SIGMOD 2015

Stop that Join! Optimizing Feature Selection over Normalized Data for Naive BayesWisconsin Database Group Seminar 2015

On Learning Generalized Linear Models over JoinsWisconsin Database Group Seminar 2014

Usability and Developability Challenges in Advanced AnalyticsIndian Institute of Technology, Madras (Invited) 2014

On Learning over JoinsMicrosoft Big Data Security Symposium (Invited) 2014Microsoft Jim Gray Systems Lab 2014

On Integrating Advanced Analytics with Scalable Structured Data ManagementWisconsin CS Preliminary Exam 2014

Scalable and Distributed PCA on REEFMicrosoft Cloud and Information Systems Lab 2013

Commoditizing Large-Scale Analytics for the Enterprise around RMicrosoft Jim Gray Systems Lab (Invited) 2013

Columbus: Feature Selection on Data Analytics SystemsWisconsin Database Group Seminar 2013

Brainwash: A Data System for Feature EngineeringCIDR 2013

Probabilistic Management of OCR Data Using an RDBMSVLDB 2012Wisconsin Database Group Seminar 2012

Large-Scale Low-Rank Matrix Factorization using Incremental Gradient DescentOracle Labs 2012

Towards a Unified Architecture for in-RDBMS AnalyticsACM SIGMOD 2012

Staccato: Probabilistic Management of OCR Data Using an RDBMSWisconsin DB Affiliates Meeting 2011

Scalable Cross-validation and Ensemble Learning in SystemMLIBM Almaden Research Center 2011

Managing Uncertainty in OCR and Speech Data Using an RDBMSMicrosoft Jim Gray Systems Lab 2011

TECHNICALSKILLS

Languages: C/C++, Java, Perl, Python, R, SQL

Data Platforms: Greenplum, Hadoop, Hive, Oracle, PostgreSQL, Spark

Hao Su

CONTACT

Address: 9500 Gilman Drive, La Jolla, CA 92093

Phone: (609) 933-0361

E-mail: haosu@eng.ucsd.edu

EDUCATION

Stanford University, Ph.D, Computer Science, USA, 2018

Beihang University, Ph.D, Mathematics, Beijing, China, 2014

Beihang University, B.S, Computer Science, China, 2006

EMPLOYMENT

University of California, San Diego, Assistant Professor of Computer Science, July, 2017 − Present

PUBLICATIONS

Referred Conferences and Journals (Representative Papers)

• Minhyuk Sung, Hao Su, Ronald Yu, Leonidas Guibas. Deep Functional Dictionaries: Learning

Consistent Semantic Structures on 3D Models from Functions. Neural Information Processing

Systems Conference (NIPS), 2018, Montreal, Canada

• Li Yi, Haibin Huang, Difan Liu, Evangelos Dalogerakis, Hao Su, Leonidas Guibas. Deep Part

Induction from Articulated Object Pairs. Transaction on Graphics (SIGGRAPH Asia), 2018,

Tokyo, Japan

• Hao Zhu, Hao Su, Peng Wang, Xun Cao, Ruigang Yang. View Extrapolation of Human

Body from a Single Image. Computer Vision and Pattern Recognition (CVPR), 2018, Salt

Lake City, USA

• Cewu Lu, Hao Su, Yongyi Lu, Li Yi, Chikeung Tang, Leonidas Guibas. Beyond Holistic

Object Recognition: Enriching Image Understanding with Part States. Computer Vision and

Pattern Recognition (CVPR), 2018, Salt Lake City, USA

• Chuang Gan, Boqing Gong, Kun Liu, Hao Su, Leonidas Guibas. Geometry-Guided CNN

for Self-supervised Video Representation learning. Computer Vision and Pattern Recognition

(CVPR), 2018, Salt Lake City, USA

• Charles R. Qi, Wei Liu, Chenxia Wu, Hao Su, Leonidas Guibas. Frustum PointNets for 3D

Object Detection from RGB-D Data. Computer Vision and Pattern Recognition (CVPR),

2018, Salt Lake City, USA

• Lin Shao, Angel X. Chang, Hao Su, Manolis Savva, Leonidas Guibas. Cross-modal Attribute

Transfer for Rescaling 3D Models. International Conference on 3D Vision, 2017, Qingdao,

China

• Charles R. Qi, Li Yi, Hao Su, Leonidas J. Guibas. PointNet++: Deep Hierarchical Feature

Learning on Point Sets in a Metric Space. Neural Information Processing Systems Conference

(NIPS), 2017, Long Beach, USA

Hao Su Page 2

• Minhyuk Sung, Hao Su, Vladimir G. Kim, Siddhartha Chaudhuri, and Leonidas Guibas. Com-

plementMe: Weakly-Supervised Component Suggestions for 3D Modeling. Transactions on

Graphics (SIGGRAPH Asia), 2017, Kyoto, Japan

• Li Yi, Leonidas J. Guibas, Aaron Hertzmann, Vladimir G. Kim, Hao Su, Ersin Yumer. Learn-

ing Hierarchical Shape Segmentation and Labeling from Online Repositories. Transactions on

Graphics (SIGGRAPH), 2017, Los Angeles, USA

• Hao Su, H. Fan, L.J.Guibas. A Point Set Generation Network for 3D Object Reconstruction

from a Single Image. Computer Vision and Pattern Recognition (CVPR), 2017, Hawaii, USA,

oral

• Hao Su, C. Qi, L.J.Guibas. PointNet: Deep Learning on Point Sets for 3D Classification and

Segmentation. Computer Vision and Pattern Recognition (CVPR), 2017, Hawaii, USA, oral

• L. Yi, Hao Su, X. Guo, L.J.Guibas. SyncSpecCNN: Synchronized Spectral CNN for 3D

Shape Segmentation. Computer Vision and Pattern Recognition (CVPR), 2017, Hawaii, USA,

spotlight oral

• S. Tulsiani, Hao Su, L. J. Guibas, A. A. Efros, J. Malik. Learning Shape Abstractions by

Assembling Volumetric Primitives. Computer Vision and Pattern Recognition (CVPR), 2017,

Hawaii, USA

• J. Shi, Y. Dong, Hao Su, S. Yu. Learning Non-Lambertian Object Intrinsics across ShapeNet

Categories. Computer Vision and Pattern Recognition (CVPR), 2017, Hawaii, USA

• K. Xu, Y. Shi, L. Zheng, J. Zhang, M. Liu, H. Huang, Hao Su, D. Cohen-Or, B. Chen.

3D Attention-Driven Depth Acquisition for Object Identification. Transactions on Graphics

(SIGGRAPH Asia), 2016

• T. Wang, Hao Su, Q. Huang, J. Huang, L. Guibas, N. Mitra. Unsupervised Texture Transfer

from Images to Model Collections. Transactions on Graphics (SIGGRAPH Asia), 2016

• L. Yi, V. Kim, D. Ceylan, I. Shen, M. Yan, Hao Su, C. Lu, Q. Huang, A. Sheffer, L. Guibas.

A Scalable Active Framework for Region Annotation in 3D Shape Collections. Transactions

on Graphics (SIGGRAPH Asia), 2016

• Hao Su, C. Qi, M. Niessner, A.Dai, M. Yan, L.J.Guibas. Volumetric and Multi-View CNNs

for Object Classification on 3D Data. Computer Vision and Pattern Recognition (CVPR),

2016, Las Vegas, USA, spotlight oral

• X. Liu, X. Fan, C. Deng, Hao Su, D. Tao. Multilinear Hyperplane Hashing. Computer Vision

and Pattern Recognition (CVPR), 2016, Las Vegas, USA

• W. Chen, H. Wang, Y. Li, Hao Su, Z. Wang, C. Tu, D. Lischinski, D. Cohen-Or, B. Chen.

Synthesizing Training Images for Boosting Human 3D Pose Estimation. International Confer-

ence of 3D Vision (3DV), 2016, Stanford, USA, oral

• Y. Xiang, W. Kim, W. Chen, J. Ji, C. Choy, Hao Su, R. Mottaghi, L. Guibas, S. Savarese.

ObjectNet3D: A Large Scale Database for 3D Object Recognition. European Conference on

Computer Vision (ECCV), 2016, Amsterdam, Netherlands, spotlight oral

• Y. Li, S. Pirk, Hao Su, C. Qi, L. Guibas. FPNN: Field Probing Neural Networks for 3D

Data. Neural Information Processing Systems Conference (NIPS), 2016, Barcelona, Spain

• A. Chang, T. Funkhouser, L. Guibas, P. Hanrahan, Q. Huang, Z. Li, S. Savarese, M. Savva, S.

Song, Hao Su, J. Xiao, L. Yi, F. Yu. ShapeNet: An Information-Rich 3D Model Repository.

arXiv preprint arXiv:1512.03012, corresponding author, 2016

Hao Su Page 3

• Hao Su, C. Qi, Y. Li, L.J.Guibas. Render for CNN: Viewpoint Estimation in Images Using

CNNs Trained with Rendered 3D Model Views. International Conference of Computer Vision

(ICCV), 2015, Santiago, Chile, oral, acceptance rate: 2%

• Hao Su, F. Wang, L. Yi, L.J.Guibas. 3D-Assisted Image Feature Synthesis for Novel Views

of an Object. International Conference of Computer Vision (ICCV), 2015, Santiago, Chile,

oral, acceptance rate: 2%

• Hao Su, Yangyan Li, Charles Qi, Noa Fish, Daniel Cohen-Or, Leonidas Guibas. Learning

Joint Embedding of Shapes and Images via CNN Image Purification. Transactions on Graphics

(SIGGRAPH Asia), 2015.

• O. Russakovsky, J. Deng, H. Su, J. Krause, S. Satheesh, S. Ma, Z. Huang, A. Karpathy,

A. Khosla, M. Bernstein, A. C. Berg, Li Fei-Fei. ImageNet: Large Scale Visual Recognition

Challenge. International Journal of Computer Vision (IJCV), 2014.

• Hao Su, Q. Huang, N. J. Mitra, Y. Li, L.J.Guibas. Estimating Image Depth using Shape

Collections. Transactions on Graphics (SIGGRAPH), 2014

• Adams Wei Yu, N. Mamoulis, H. Su. Reverse Top-k Search using Random Walk with Restart.

Very Large Database (VLDB), 2014, China. full oral

• Hao Su, A.W. Yu, L. Fei-Fei. Efficient Euclidean Projections onto the Intersection of Norm

Balls. International Conference on Machine Learning (ICML), 2012, Edinburgh, UK, full

oral

• Hao Su, L.-J. Li, E. P. Xing, L. Fei-Fei. Object Bank: A High-Level Image Representation

for Scene Classification & Semantic Feature Sparsification. Neural Information Processing

Systems Conference (NIPS), 2010, Vancouver, Canada.

• Hao Su, Min Sun, Li Fei-Fei, Silvio Savarese, Learning a dense multi-view representation

for detection, viewpoint classification and synthesis of object categories. In proceedings of the

International Conferences on Computer Vision and Pattern Recognition (ICCV), 2009, Kyoto,

Japan, oral, acceptance rate: 4%

• Hao Su, Min Sun, Silvio Savarese, Li Fei-Fei, A Multi-View Probabilistic Model for 3D Object

Classes. Computer Vision and Pattern Recognition (CVPR), 2009, Miami, USA

• Kun Yang, H. Su, W.H. Wong. co-BPM: a Bayesian Model for Estimating Divergence and

Distance of Distributions. arxiv.

PROFESSIONAL EXPERIENCE• Area Chair, Computer Vision and Pattern Recognition (CVPR), Long Beach, USA, 2019

• Program Chair, International Conference on 3D Vision (3DV), Qingdao, China, 2017

• Publication Chair, International Conference on 3D Vision (3DV), Stanford, USA, 2016

• Chair, Workshop on Virtual/Augmented Reality for Visual Artificial Intelligence (affiliated

with ECCV and ACM-MM), Amsterdam, Netherlands, 2016

• Chair, Workshop on 3D Representation and Recognition (affiliated with ICCV), Santiago,

Chile, 2015

• Chair, Workshop on 3D from Single Images (affiliated with CVPR), Boston, USA, 2015

• Organizer, Tutorial on 3D Deep Learning (affliated with CVPR), Hawaii, USA, 2017

• Organizer, Large-scale 3D Shape Retrieval Challenge, a subtrack of SHREC2016 (affiliated

with EUROGRAPHICS), Lisbon, Portugal, 2016

Hao Su Page 4

• Program Committee, First Workshop on Virtual Reality meets Physical Reality: Modelling

and Simulating Virtual Humans and Environments (affiliated with SIGGRAPH Asia) Macau,

2016

• Program Committee, CVPR, 2016-2017

• Program Committee, ICCV, 2017

• Reviewer of CVPR, 2010-2017

• Reviewer of ICCV, 2009-2015

• Reviewer of SIGGRAPH, 2014-2017

• Reviewer of SIGGRAPH Asia, 2015-2016

• Reviewer of ICRA, 2017

• Reviewer of EUROGRAPHICS (EG), 2015,2016

• Reviewer of NIPS, 2011,2016

• Reviewer of PACIFIC GRAPHICS (PG), 2014

• Reviewer of Computer Graphics Forum

• Reviewer of Transactions on Graphics (TOG)

• Reviewer of Transactions on Pattern Analysis and Machine Intelligence (TPAMI)

• Reviewer of Transactions on Information Processing (TIP)

INVITED TALKS• Workshop on Autonomous Driving, ECCV, 2018

• GRASP Lab, University of Pennsylvania, 2018

• Sumsung Research, USA, 2017

• Google X, Mountain View, USA, 2017

• Shape, Solid, Structure, & Physical Modeling (S3PM), 2017

• University of California, San Diego, 2017

• Cornell University, 2017

• NVIDIA, USA, 2017

• Workshop on 3D Deep Learning, NIPS, 2016

• CSAIL & CogSci, MIT, USA, 2016

• Workshop on Understanding 3D and Visuo-Motor Learning, 3DV, 2016

• Facebook, USA, 2016

• Adobe Research, USA, 2016

• Stanford CS Faculty Meeting, Stanford, USA, 2016

• University of California, Berkeley, USA, 2016

• Johannes Gutenberg University, Germany, 2015

• Max-Planck-Institute (MPI), Germany, 2015

• Tsinghua University, China, 2015

• Shandong University, China, 2015

Hao Su Page 5

• Beihang University, China, 2015

• University of California, Berkeley, 2015

• University of California, Los Angeles, 2015

1

DR. JADE D’ALPOIM GUEDES Scripps Institution of Oceanography-Sverdrup Hall

8615 Kennel Way Rm # 1255 La Jolla, CA 92037-0212

Tel: +857 600 6485; Email: jguedes@ucsd.edu

www.jadeguedes.com (a) Professional Preparation INALCO Paris, Tibetan & Himalayan Studies (honors) DULCO 2002 Paris IV-Sorbonne Art History & Archaeology (honors) Licence 2004 Paris VII Denis Diderot Sinology (honors) DEUG 2005 Harvard University Anthropology (Archaeology) Ph.D 2013 Harvard University Earth and Planetary Science Postdoctoral Fellow 2014 (b) Appointments 2017-Present Assistant Professor, Scripps Institution of Oceanography, Department of

Anthropology, University of California, San Diego 2016-Present Research Associate, Department of Organismic and Evolutionary

Biology, Harvard University 1/2014-2017 Assistant Professor, Department of Anthropology, Washington State

University 1/2014-Present Visiting Scholar, Sichuan Provincial Institute of Archaeology 01/2012-06/2012 Teaching Fellow, Department of Organismal and Evolutionary Biology,

Harvard University 01/2010-08/2013 Instructor of Record, Sichuan University, China 01/2009-06/2010 Head Teaching Fellow, Department of Anthropology, Harvard University,

(Harvard Derek Bok Center Award for Excellence in Teaching) 2008-Present Analyst for Macrobotanical Remains, Chengdu City Institute of

Archaeology 2007-2011 Resident Tutor in Social Sciences, Adams House, Harvard College (c) Five Relevant Products

(i) Five publications most closely related to proposed project

J. d’Alpoim Guedes, R. Kyle Bocinsky, Stefani Crabtree & Timothy Kohler (2016) 21st Century Approaches to Ancient Problems: Climate and Society. Proceedings of the National Academy of the Sciences 113(51): 14483-14491

J. d’Alpoim Guedes, Sturt Manning, R. Kyle Bocinsky (2016). A 5500 year model of changing crop niches on the Tibetan Plateau. Current Anthropology. 57(4): 517-522.

J. d’Alpoim Guedes, Lu Hongliang, Anke Hein & Amanda Schmidt (2015). Early Evidence for the use of wheat and barley as staple crops on the margins of the Tibetan Plateau. Proceedings of the National Academy of the Sciences 112(8): 5625-5630.

J. d’Alpoim Guedes, Kyle Bocinsky & Ethan Butler(2015). Comment on “Agriculture facilitated permanent human occupation of the Tibetan Plateau after 3600 B.P” Science 348 (6237): 872. J. d'Alpoim Guedes, H. Lu, Y. Li, R. Spengler, X. Wu, M. Aldenderfer (2013) Early Agriculture on the Tibetan Plateau: The Archaeobotanical Evidence Archaeological and Anthropological Sciences (DOI) 10.1007/s12520-013-0153-4

(ii) Five other significant contributions J. d’Alpoim Guedes, Jin Guiyun, R. Kyle Bocinsky (2015) The Impact of Climate on the Spread of Rice Agriculture to North-Eastern China: An Example from Shandong. PLOS-One 10(6): e013043.

2

J. M Marston, J. d'Alpoim Guedes and C. Warinner (eds.) (2014) Method and Theory in Paleoethnobotany. University Press of Colorado, Boulder. J. d’Alpoim Guedes (2015). Rethinking the Spread of Agriculture to the Tibetan Plateau. The Holocene doi: 10.1177/0959683615585835

J. d’Alpoim Guedes, M. Jiang, K. He, Z. Jiang, X. Wu (2013) New Evidence for the Spread of Early Rice and Foxtail Millet Agriculture in Southwest China Antiquity 87 (337): 758-771. J. d’Alpoim Guedes, D. Carrasco, R. Flad, E. Fosse, M. Herzfeld, C. C. Lamberg-Karlovsky, C. M. Lewis, M. Liebmann, R. Meadow, N. Patterson, M. Price, M. Reiches, S. Richardson, H. Shattuck-Heidorn, J. Ur, G. Urton, C. Warinner (2013) Is Poverty in our Genes?: A Reply to Ashraf and Galor. Current Anthropology 54(1): 71-79. (d) Synergistic Activities Co-Founder (with Dr. Christina Warinner) www.paleobot.org, a website that provides a platform for researchers to share reference collection images and engage in collaborative identification of unknown archaeobotanical specimens, share identification guides, and for researchers to upload profiles that facilitate communication, networking, and collaboration. 1.) Public Engagement and Lectures: Over many years I have given public lectures and seminars including ““Food, Biodiversity and Climate Change: Lessons from the past for the future” at the Osher Life Long Learning Center and at the Peabody Museum. I have also led guided public tours and given lectures at the San Diego Botanical Gardens. Worked with the Heritage Grain conservancy to carry out research on landrace varieties of millet and wheat. 2.) Teaching, advising and mentoring of students in archaeology. I have taught and advised undergraduate and graduate students in the fields of archaeology, archaeological science. I have sat on 15 graduate committees between Washington State University and UCSD. Three of my students have graduated and have found positions in the field. I have also mentored several postdoctoral fellows, both of which have been awarded early faculty positions. I currently mentor seven undergraduate students in my laboratory, several of which are underrepresented minorities including two McNair fellows. I am also committed to training students globally and set up laboratories for archaeobotanical and human osteological analyses and trained students at Sichuan University, the Chengdu City Institute of Archaeology and the Sichuan Provincial Institute of Archaeology in the Peoples Republic of China. My interest in preparing graduate students for the job market has also lead me to organize a seminar for graduate students in our department entitled “Tackling the academic job market”. In this seminar, I helped graduate students plan out their trajectory from early graduate school through graduation through the creation of a five-year plan. 3.) Service to the field: I am currently the Secretary of the Society of Ethnobiology. Two years ago, I founded and have served as the President for the Association for Women in Asian Archaeology: an association that aims to promote the representation and success of women working in the archaeology of Asia. I am also on the Board of Directors for the Institute of Field Research. As part of the board of directors, I rate applications from individuals who apply to have their program become an IFR field school, carry out site visits at potential and current field schools and carry out quality control assessments. I hold two key editorial roles: I am currently the editor of the Bulletin for the Society of East Asian Archaeology. I have one other editorial role where I serve as an archaeology editor for the series Contributions in Ethnobiology: a book series organized by the Society for Ethnobiology. I am an in demand reviewer and have reviewed on average 15+ articles a year.

Julian McAuley

Computer Science DepartmentUniversity of California San Diego9500 Gilman DrLa Jolla, CA 92093 Phone: 650 521 3166

email: jmcauley@eng.ucsd.eduurl: http://cseweb.ucsd.edu/~jmcauley/

Current positionAssistant Professor, Computer Science Department, University of California San DiegoSpecializing in Machine Learning • Data Mining • Recommender Systems

Education2011-2014 Postdoctoral Scholar, Stanford University

Advised by Jure Leskovec

2008-2011 PhD, Australian National UniversityAdvised by Tibério CaetanoThesis: Graphical Models for Inference and Learning in Computer Vision

2003-2007 BEng, BSc, University of New South WalesSoftware Engineering and Pure Mathematics (first-class honours and the University Medal)

Professional experience2010 Intern, Google2009 Intern, Xerox Research Centre Europe2006-2008 Researcher, NICTA (part time)

Selected Publications2016 McAuley, J. and Yang, A. (2016), “Addressing complex and subjective product-related queries with

customer reviews,”WWW

2016 He, R. and McAuley, J. (2016) “Ups and downs: Modeling the visual evolution of fashion trendswith one-class collaborative filtering,”WWW

2016 He, R. andMcAuley, J. (2016) “VBPR: Visual bayesian personalized ranking from implicit feedback,”AAAI

2015 McAuley, J., Pandey, R. and Leskovec, J. (2015) “Inferring networks of substitutable and comple-mentary products,” KDD

2013 McAuley, J. and Leskovec, J. (2013), “Hidden factors and hidden topics: understanding rating di-mensions with review text,” RecSys

1

2013 McAuley, J. and Leskovec, J. (2013), “From amateurs to connoisseurs: modeling the evolution ofuser expertise through online reviews,”WWW

2012 McAuley, J. and Leskovec, J. (2012), “Learning to discover social circles in ego networks,” NIPS

For a complete list, including workshop papers, book chapters, and patents, see my homepage andGoogle Scholar profile.

FundingExtramural Funding

2019 Samsung . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . $100,0002018 Adobe (Digital Marketing Research Award) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . $50,0002018 Flipkart . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . $25,0002018-2023 National Science Foundation (CAREER Award NSF-IIS-1750063) . . . . . . . . . . . . . . . . . . . . . . . $550,000

“Structured output models of recommendations, activities, and behavior”2017 Amazon . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . $80,0002016-2018 National Science Foundation (NSF-IIS-1636879) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . $517,817

“Knowledge discovery and real-time interventions from sensory data flows in urban spaces”(with Altintas, I., Gadh, R., Gupta, R., Shutters, S., and Srivastava, M.)

2016 Symantec . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . $30,0002016-2019 Australian Research Council (DP-160100703) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . aud$318,000

“Probabilistic graphical models for interventional queries” (with Shi, Q. and M. Pawan Kumar)2015-2017 Adobe (four funding rounds) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . $32,0002015, 2016 NVIDIA (two funding rounds) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . hardware donation2015 Kosei . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . $30,000

Intramural Funding

2018 Chancellor’s Research Excellence Scholarship (with Puckette, M.) . . . . . . . . . . . . . . . . . . . . . . $28,0002018 Chancellor’s Interdisciplinary Collaboratories . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . $25,750

(with Hsu, C.-N., Nakashole, N., and Viirre, E.)2017-2018 Frontiers of Innovation Scholars Program . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . $50,000

(two projects, with Moore, D., Brouwer, K., Burr, A., and Tolley, C.)2016 “MetroLab” (City of San Diego / UCSD Office of the Chancellor) . . . . . . . . . . . . . . . . . . . . . . . . . $45,000

(with Altintas, I. and Krstic, M.)

Reviewing & ServiceAssociate Editor: Pattern Recognition.

Program Committees: AAAI, AISTATS, CIKM, EMNLP, ICML, ICWSM, IJCAI, KDD, RecSys, SDM,SIGIR, UAI, WSDM, and WWW.

Other Reviewing: JMLR, PAMI, TIST, TKDD, TKDE, TWEB, CVPR, NIPS, SIGMOD, and VLDB.

Chair: Workshop on Mining and Learning with Graphs (2013, 2016, 2017); Workshop on Infer-ence in Graphical Models with Structured Potentials (2011); Workshop on Big Graphs, Theory andPractice (2016); Southern California Machine Learning Symposium (2016); Workshop on MachineLearning meets Fashion (2017, 2018).

2

Leo Porter

leporter@eng.ucsd.eduDepartment of Computer Science and Engineering

9500 Gilman DriveLa Jolla, CA 92093

http://cseweb.ucsd.edu/∼leporter/

Education

Doctorate of Philosophy in Computer ScienceUniversity of California, San Diego. 2011.Advisor: Dean M. TullsenThesis Title: Single Thread Performance in the Multi-Core Era

Master of Science in Computer ScienceUniversity of California, San Diego. 2007.Advisor: Dean M. Tullsen

Bachelor of Arts in Computer ScienceUniversity of San Diego. 2000.Graduated magna cum laude with departmental honors

Research Interests

Computer Science Education: peer instruction, flipped classrooms, active learning, conceptinventories, media computation, assessment, and retention.

Computer Architecture: chip-multiprocessors, simultaneous multithreading, speculativemultithreading, transactional memory, thread-level parallelism, branch prediction, cache design,and architecture-aware scheduling.

Professional Employment History

Associate Teaching Professor, UC San Diego, 2018–Present

Assistant Teaching Professor, UC San Diego, 2014–2018

Assistant Professor, Skidmore College, 2011–2014

Consultant, EP Analytics, 2013–2014

Officer, US Navy, 2000–2004

Computer Science Education Publications

• Soohyun Nam Liao, William G. Griswold, Leo Porter. Classroom Experience Report on JigsawLearning. In Proceedings of the Conference on Innovation and Technology in Computer ScienceEducation, July, 2018.

• David P. Bunde, Cynthia Taylor, Jaime Spacco, Andrew Petersen, Soohyun Nam Liao, Leo Porter.A Multi-institution Exploration of Peer Instruction in Practice. In Proceedings of the Conferenceon Innovation and Technology in Computer Science Education, July, 2018.

• Leo Porter, Daniel Zingaro, Cynthia Lee, Cynthia Taylor, Kevin C. Webb, Michael Clancy.Developing Course-Level Learning Goals for Basic Data Structures in CS2. In Proceedings of theSpecial Interest Group on Computer Science Education (SIGCSE) Technical Symposium, February,2018.

• Daniel Zingaro, Michelle Craig, Leo Porter, Brett A. Becker, Yingjun Cao, Phill Conrad, DianaCukierman, Arto Hellas, Dastyni Loksa, Neena Thota. Achievement Goals in CS1: Replication andExtension. In Proceedings of the Special Interest Group on Computer Science Education (SIGCSE)Technical Symposium, February, 2018.

• Mia Minnes, Christine Alvarado, and Leo Porter. Lightweight Techniques to Support Students inLarge Classes. In Proceedings of the Special Interest Group on Computer Science Education(SIGCSE) Technical Symposium, February, 2018.

• Yingjun Cao and Leo Porter. Impact of Performance Level and Group Composition on StudentLearning during Collaborative Exams. In Proceedings of the Conference on Innovation andTechnology in Computer Science Education, July, 2017.

• Leo Porter, Cynthia Lee, Beth Simon, and Mark Guzdial. Preparing tomorrow’s faculty to addresschallenges in teaching computer science. Communications of the ACM, 60(5), May 2017.

• Soohyun Nam Liao, William G. Griswold, and Leo Porter. Impact of Class Size on StudentEvaluations for Traditional and Peer Instruction Classrooms. In Proceedings of the Special InterestGroup on Computer Science Education (SIGCSE) Technical Symposium, March, 2017.

• Christine Alvarado, Mia Minnes and Leo Porter. Micro-Classes: A Structure for Improving StudentExperience in Large Classes. In Proceedings of the Special Interest Group on Computer ScienceEducation (SIGCSE) Technical Symposium, March, 2017.

• Yingjun Cao and Leo Porter. Evaluating Student Learning from Collaborative Group Tests inIntroductory Computing. In Proceedings of the Special Interest Group on Computer ScienceEducation (SIGCSE) Technical Symposium, March, 2017.

• Soohyun Nam Liao, Daniel Zingaro, Michael A. Laurenzano, William G. Griswold, and Leo Porter.Lightweight, Early Identification of At-Risk CS1 Students. In Proceedings of the Conference onInternational Computing Education Research, September, 2016.

• Simon, Judy Sheard, Daryl D’Souza, Peter Klemperer, Leo Porter, Juha Sorva, Martijn Stegeman,and Daniel Zingaro. Benchmarking Introductory Programming Exams: Some Preliminary Results.In Proceedings of the Conference on International Computing Education Research, September,2016.

• Yingjun Cao, Leo Porter, and Daniel Zingaro. Examining the Value of Analogies in IntroductoryComputing. In Proceedings of the Conference on International Computing Education Research,September, 2016.

• Simon, Judy Sheard, Daryl D’Souza, Peter Klemperer, Leo Porter, Juha Sorva, Martijn Stegeman,and Daniel Zingaro. Benchmarking Introductory Programming Exams: How and Why. InProceedings of the Conference on Innovation and Technology in Computer Science Education, June,2016.

• Leo Porter, Dennis Bouvier, Quintin Cutts, Scott Grissom, Cynthia Lee, Robert McCartney, DanielZingaro, and Beth Simon. A Multi-Institutional Study of Peer Instruction in IntroductoryComputing. In Proceedings of the Special Interest Group on Computer Science Education TechnicalSymposium, March, 2016.Winner – Best Paper AwardReprinted in ACM Inroads, June, 2016.

• Daniel Zingaro and Leo Porter. Impact of Student Achievement Goals on CS1 Outcomes. InProceedings of the Special Interest Group on Computer Science Education Technical Symposium,March, 2016.

• Daniel Zingaro and Leo Porter. Tracking Student Learning from Class to Exam using IsomorphicQuestions. In Proceedings of the Special Interest Group on Computer Science Education TechnicalSymposium, March, 2015.

• Cynthia Taylor, Daniel Zingaro, Leo Porter, Kevin C. Webb, Cynthia Bailey Lee, and Mike Clancy.Computer Science Concept Inventories: Past and Future. In Computer Science Education 24.4,Special Issue: Concept Inventories: 253-276, 2014.

• Leo Porter, Daniel Zingaro, and Raymond Lister. Predicting Student Success Using Fine GrainClicker Data. In Proceedings of the 10th Annual International Computing Education ResearchConference, August, 2014.Winner – Chair’s Award

• Leo Porter, Cynthia Taylor, and Kevin C. Webb. Leveraging Open Source Principles for FlexibleConcept Inventory Development. In Proceedings of the 19th Annual Conference on Innovation andTechnology in Computer Science Education, June, 2014.

• Daniel Zingaro and Leo Porter. Peer Instruction: A Link to the Exam. In Proceedings of the 19thAnnual Conference on Innovation and Technology in Computer Science Education, June, 2014.

• Leo Porter and Daniel Zingaro. Importance of Early Performance in CS1: Two ConflictingAssessment Stories. In Proceedings of the Special Interest Group on Computer Science EducationTechnical Symposium, March, 2014.

• Daniel Zingaro and Leo Porter. Peer Instruction in Computing: The Value of InstructorIntervention. Computers and Education, Volume 71, February, 2014.

• Cynthia Bailey Lee, Saturnino Garcia, and Leo Porter. Can Peer Instruction Be Effective inUpper-Division Computer Science Courses? ACM Transactions on Computing Education SpecialIssue on Alternatives to Lecture, August, 2013.

• Beth Simon, Sarah Esper, Leo Porter, and Quintin Cutts. Student Experience in aStudent-Centered Peer Instruction Classroom. In Proceedings of the 9th Annual InternationalComputing Education Research Workshop, August, 2013.

• Leo Porter, Mark Guzdial, Charlie McDowell, and Beth Simon. Success in IntroductoryProgramming: What Works? Communications of the ACM, 58(8),34-36, August, 2013.

• Leo Porter, Saturnino Garcia, Hung-wei Tseng, and Daniel Zingaro. Evaluating StudentUnderstanding of Core Concepts in Computer Architecture. In Proceedings of the 18th AnnualConference on Innovation and Technology in Computer Science Education, July, 2013.

• Leo Porter, Saturino Garcia, John Glick, Andrew Matusiewicz, and Cynthia Taylor. PeerInstruction in Computer Science at Small Liberal Arts Colleges. In Proceedings of the 18th AnnualConference on Innovation and Technology in Computer Science Education, July, 2013.

• Leo Porter and Beth Simon. Retaining Nearly One-Third more Majors with a Trio of InstructionalBest Practices in CS1. In Proceedings of the Special Interest Group on Computer ScienceEducation Technical Symposium, March, 2013.Winner – Best Paper Award

• Leo Porter, Cynthia Bailey Lee, and Beth Simon. Halving Fail Rates using Peer Instruction: AStudy of Four Computer Science Courses. In Proceedings of the Special Interest Group onComputer Science Education Technical Symposium, March, 2013.

• Daniel Zingaro, Cynthia Bailey Lee, and Leo Porter. Peer Instruction in Computing: the Role ofReading Quizzes. In Proceedings of the Special Interest Group on Computer Science EducationTechnical Symposium, March, 2013.

• Leo Porter and Beth Simon. Fostering Creativity in CS1 by hosting a Computer Science Art Show.ACM Inroads, March, 2013.

• Leo Porter, Cynthia Bailey-Lee, Beth Simon, and Daniel Zingaro. Peer Instruction: Do StudentsReally Learn from Peer Discussion in Computing? In The 7th Annual International ComputingEducation Research Workshop, August, 2011.

• Leo Porter, Cynthia Bailey-Lee, Beth Simon, Quintin Cutts, and Daniel Zingaro. ExperienceReport: A Multi-classroom Report on the Value of Peer Instruction. In The 16th AnnualConference on Innovation and Technology in Computer Science Education, June, 2011.

• Beth Simon, Paivi Kinnunen, Leo Porter, and Dov Zazkis. Experience Report: CS1 for Majors withMedia Computation. In The Fifteenth Annual Conference on Innovation and Technology inComputer Science Education, June, 2010.

Computer Architecture Publications

• Craig Disselkoen, David Kohlbrenner, Leo Porter, and Dean Tullsen. Prime+Abort: A Timer-FreeHigh-Precision L3 Cache Attack using Intel TSX. In Proceedings of USENIX Security 2017,August, 2017.

• Leo Porter, Michael A. Laurenzano, Ananta Tiwari, Adam Jundt, William A. Ward, Jr., RoyCampbell, and Laura Carrington. Making the Most of SMT in HPC: System- and Application-LevelPerspectives. In ACM Transactions on Architecture and Code Optimization 11.4, January, 2015.Presented at HiPEAC 2015.

• Alex D. Breslow, Leo Porter, Ananta Tiwari, Michael Laurenzano, Laura Carrington, Dean M.Tullsen, and Allan E. Snavely. The Case for Colocation of HPC Workloads. Concurrency andComputation: Practice and Experience. Special issue on the Analysis of Performance and Power forHighly Parallel Systems, 2013.

• Jeff Brown, Leo Porter, and Dean M. Tullsen. Fast Thread Migration via Cache Working SetPrediction. In Seventeenth International Symposium on High Performance Computer Architecture,February, 2011.Winner – Best Student Paper Award

• Leo Porter, Bumyong Choi, and Dean M. Tullsen. Mapping Out a Path from HardwareTransactional Memory to Speculative Multithreading. In Eighteenth International Conference onParallel Architectures and Compilation Techniques, September, 2009.

• Leo Porter and Dean M. Tullsen. Creating Artificial Global History to Improve Branch PredictionAccuracy. In Twenty-Third International Conference on Supercomputing, June, 2009.

• Bumyong Choi, Leo Porter, and Dean M. Tullsen. Accurate Branch Prediction for Short Threads.In Thirteenth International Conference on Architectural Support for Programming Languages andOperating Systems, March, 2008.

Research Grants

• NSF IUSE, Identifying and Aiding At-Risk Students in Computing, Principal Investigator,$299,326, 2017-2020

• NSF IUSE, Collaborative Research: Infrastructure and Development of a Computer ScienceConcept Inventory for CS2, Principal Investigator, $125,508, 2015-2018

• Coursera, Intermediate Software Specialization, Principal Investigator, $300,000, 2015-2018

• NSF IUSE, EAGER: The Experience of a Small College, at a Large Scale, Co-PrincipalInvestigator, $300,000, 2015-2016

• NSF IUSE, Collaborative Research: A New Computer Science Faculty Teaching Workshop,Co-Principal Investigator, $33,040, 2015-2018

• NSF TUES, Collaborative Research: Peer Instruction in Computer Science, Co-PrincipalInvestigator, $57,818, 2012-2015

Teaching Experience

Associate Teaching Professor, UC San Diego, July 2014–Present

• CSE8B - Introduction to Computer Science in Java (2 quarters)• CSE100 - Advanced Data Structures (2 quarters)• CSE141 - Introduction to Computer Architecture (7 quarters)• CSE240A - Graduate Computer Architecture (3 quarters)• CSE599 - Teaching Methods in Computer Science (4 quarters)• On average, 97% of the students recommended the instructor on course teaching evaluations.• (COSMOS Summer Program) Instructor - Computers in Everyday Life (2 summers)• (Coursera) Java Programming: Object-Oriented Design of Data Structures Specialization

∗ Course 1: Object Oriented Programming in Java∗ Course 2: Data Structures: Measuring and Optimizing Performance∗ Course 3: Advanced Data Structures in Java∗ Course 4: Mastering the Software Engineering Interview∗ Capstone: Analyzing (Social) Network Data∗ Over 250,000 enrolled learners across specialization. Average Course Rating of 4.7/5.0.

• (edX MicroMasters) DSE200x: Python for Data Science∗ Launched June 2017, over 125,000 enrolled learners.

Assistant Professor, Skidmore College, September 2011–July 2014

• CS106 - Computer Science I (6 semesters)• CS318 - Computer Organization and Design (3 semesters)• CS376B - Advanced Computer Architecture• CS376B - Operating Systems• HF-200 - Honors Forum Topic: Plagues and Peoples• HF-200 - Honors Forum Topic: Genetically Modified Organisms• Average instructor rating was 4.6 out of 5. College mean for all instructors was 4.2.

Summer Graduate Teaching Fellow (SGTF), UC San Diego, August–September 2010

• CSE 141 - Introduction to Computer Architecture• CSE 141L - Project in Computer Architecture

Adjunct Professor, University of San Diego, January-May 2009

• COMP300 - Digital Hardware and Digital Hardware Lab

Student Research/Independent Study

UC San Diego

Ph.D. Thesis Co-advisor, Sander Valstar, Fall 2017–Present.

Ph.D. Thesis Co-advisor, Soohyun Nam Liao, Fall 2015–Present.

U/G Honors Thesis, Marshall Seid, Fall 2016–Spring 2017.

MS Project, Brandon Williams, Spring 2016.

CSE198, Student Independent Study, Winter 2015.

Skidmore College

Nikolay Yosifov, Resource Allocation in SMT Processors, Summer 2013.

Samuel Gunther, Senior Thesis - Resource Partitioning for SMT Processors, 2012-2013.

Pierre-Francois Wolfe, Stenography Techniques for Message Hiding, Spring 2013

Canaan Gifford, Virtual Memory Simulation, Fall 2012

Adina Micula and Haoran Ma, Prefetching for SpMT Processors, Summer 2012.

• Presented at Grace Hopper Celebration of Women in Computer, October 2013.

Samuel Gunther, Impact of History Length Tuning on Branch Prediction, Spring 2012

Sarah Llewelyn, Hardware Techniques to Counter Buffer Overflow Attacks, Fall 2011

Invited Talks

Breaking Boundaries in STEM Education Conference, April 2017

• Computational Thinking: Computing Across Domains

COSMOS Summer Program, UC San Diego, July 2016

• Do “Moore” Cores Lead to Better Performance

CAMSEE, UC San Diego, April 2016

• Coursera MOOC Materal Development and Impacts

CAMSEE, August 2015

• Collaborative Research Projects in CSE

CCSC Eastern Region, York College, October 2014

• Predicting Student Success Using Fine Grain Clicker Data

Union College, Ocober 2013

• Single Thread Performance in the Multi-core Era

Skidmore College Community Lectures, October 2013

• The Future of Computing: Do “Moore” Transistors Still Mean Faster Computers?

Oberlin College, April 2013

• Single Thread Performance in the Multi-core Era

Skidmore College Faculty Summit, January 2013

• Elearning

Skidmore College Faculty to Faculty Series, November 2012

• Peer Instruction: Active Learning in Computer Science

Skidmore College Mature Learners Program, November 2012

• The Multi-core Era: Do “Moore” Transistors Still Mean Faster Computing?

SUNY Potsdam, September 2012

• Single Thread Performance in the Multi-core Era

Skidmore College Leadership Program, July 2012

• Innovation at Skidmore College

NERCOMP SIG: Back to the Clicker, April 2012

• Peer Instruction: Do Students Really Learn from Peer Discussion in Computing?

ICER Lightning Talks, August 2011

• Concept Inventories in Computer Architecture

University of San Diego, April 2009

• Accurate Branch Prediction for Short Threads

University Service

UC San Diego

Chair — Masters Curriculum Committee, 2017–Present

Chair — CSE Academic Integrity Committee, 2016–Present

Representative — CSE Representative to the Academic Senate, 2015–2017

Member — CSE Undergraduate Curriculum Committee, 2014–2017

Member — Center for Advancing Mathematics, Science, and Engineering Education, 2014–Present

Skidmore College

Member — Student Affairs Subcommittee, 2012–2014

Director — Bi-Annual Computer Science Art Show, 2011–2014

Director — Bi-Annual Critters Tournament, 2011–2014

Professional Service

Conference Chair, CCSC Southwest Region, 2017

Co-Host, New Computer Science Faculty Teaching Workshop, 2015–Present

Associate Editor, Transactions on Computing Education, 2016–Present

Co-Editor, SIGCSE Bulletin, 2016–Present

PC Member and Lightning Talks/Poster Chair, ICER, 2015, 2016

PC Member and Workshop Chair, CCSC Southwest Region, 2015, 2016

Workshop Co-Host, Evidence-Based Teaching Practices in CS, SIGCSE 2017

PC Member, Grace Hopper Celebration Educational Technologies, 2013

Reviewer: ICER, CSE, C&E, TOE, TOCE, ACM Inroads, SIGCSE, ITiCSE, CCSC, PLOS ONE,IEEE MICRO, ISCA, and HPCA

Examiner: Oberlin College Computer Science Departmental Honors. 2013

Hosted Peer Instruction Workshops at CCSCSW11, SIGCSE12, SIGCSE13, CCSCNE13, GHC13,SIGCSE16, and as part of the C2GEN Workshop Series in 2014 and 2015

Submission and Review Web Chair, ISPASS 2008 and IEEE MICRO Top Picks 2008

Senior Member, ACM

Military Service and Awards

Lieutenant, Honorable Discharge, United States Navy, 2008

Navigator, USS Milius (DDG-69), San Diego, CA, 2003–2004

Auxiliary and Electrical Officer, USS Milius (DDG-69), San Diego, CA, 2001–2003

Veteran - Operation Iraqi Freedom

Henry O’Reilly Graduate Scholarship from the Reserve Officer Association: 2005, 2006, 2007, 2008

Awarded Three Navy Achievement Medals: 2002, 2003, 2004

References

Available on Request

MANMOHAN CHANDRAKER

Computer Science and EngineeringUniversity of California, San DiegoLa Jolla, CA 92093

Phone: (858) 401-0407Email: mkchandraker@eng.ucsd.eduWeb: cseweb.ucsd.edu/~mkchandraker

EDUCATION

2003 – 2009 University of California, San DiegoPh.D. in Computer ScienceThesis: From Pictures to 3D: Global Optimization for Scene Reconstruction

1999 – 2003 Indian Institute of Technology, BombayB.Tech. in Electrical Engineering

WORK EXPERIENCE

2016 – Present University of California, San Diego Assistant Professor2015 – 2016 NEC Labs America Department Head2014 – 2015 NEC Labs America Senior Researcher2011 –2014 NEC Labs America Researcher2009 – 2011 University of California, Berkeley Postdoctoral Scholar

HONORS

2018 NSF CAREER Award2016 Outstanding Reviewer Award at ECCV2014 IEEE Computer Society Best Paper Award at CVPR2011 IEEE PAMI Special Issue on Best Papers of CVPR 20112010 UCSD nominee for ACM Doctoral Dissertation Award2009 IEEE Computer Society Outstanding Reviewer Award at ICCV2009 CSE Dissertation Award for Best Thesis at UCSD2007 Marr Prize Honorable Mention for Best Paper at ICCV

PUBLICATIONS

Journal Articles ††

1. C. Li, Z. Zia, Q.-H. Tran, X. Yu, G. Hager and M. Chandraker. Deep Supervision withIntermediate Concepts. IEEE Transactions on Pattern Analysis and Machine Intelligence, PAMI.[to appear]

2. T.-C. Wang, M. Chandraker, A. Efros and R. Ramamoorthi. SVBRDF-Invariant Shape andReflectance Recovery from Light Fields. IEEE Transactions on Pattern Analysis and MachineIntelligence, PAMI 40(3):740-754, March 2018.

3. M. Chandraker. The Information Available to a Moving Observer on Shape Recovery withUnknown Isotropic BRDF. IEEE Transactions on Pattern Analysis and Machine Intelligence,PAMI 38(7):1283-1297, July 2016. [Special Issue, Best Papers of CVPR 2014]

4. M. Chandraker, J. Bai and R. Ramamoorthi. On Differential Photometric Reconstruction withUnknown, Isotropic BRDFs. IEEE Transactions on Pattern Analysis and Machine Intelligence,PAMI 35(12):2941-2955, December 2013 [Special Issue, Best Papers of CVPR 2011].

5. M. Chandraker, J. Bai, T.-T Ng and R. Ramamoorthi. On the Duality of Forward andInverse Light Transport. IEEE Transactions on Pattern Analysis and Machine Intelligence, PAMI33(10):2122-2128, October 2011.

††IEEE PAMI and IJCV have among the highest ISI impact factors across all computer science categories.

6. M. Chandraker, S. Agarwal, D.J. Kriegman and S. Belongie. Globally Optimal Algorithmsfor Stratified Autocalibration. International Journal of Computer Vision, IJCV 90(2):236-254,November 2010. [Special Issue, Best Papers of ICCV 2007]

7. F. Kahl, S. Agarwal, M. Chandraker, D.J. Kriegman and S. Belongie. Practical GlobalOptimization for Multiview Geometry. International Journal of Computer Vision, IJCV 79(3):271-284, September 2008.

Refereed Conferences ‡‡

8. Z. Li, Z. Xu, R. Ramamoorthi, K. Sunkavalli and M. Chandraker. Learning to ReconstructShape and Spatially-Varying Reflectance from a Single Image. Siggraph Asia 2018. [to appear]

9. Z. Li, K. Sunkavalli and M. Chandraker. Materials for Masses: SVBRDF Acquisition with aSingle Mobile Phone Image. ECCV 2018. [oral presentation]

10. S. Schulter, M. Zhai, N. Jacobs and M. Chandraker. Learning to Look around Objects forTop-View Representations of Outdoor Scenes. ECCV 2018.

11. M. Fathy, Q.-H. Tran, Z. Zia, P. Vernaza and M. Chandraker. Hierarchical Metric Learningand Matching for 2D and 3D Geometric Correspondences. ECCV 2018.

12. Y.-H. Tsai, W.-C. Hung, S. Schulter, K. Sohn, M.-H. Yang and M. Chandraker. Learningto Adapt Structured Output Space for Semantic Segmentation. IEEE Conference on ComputerVision and Pattern Recognition, CVPR 2018.

13. Z. Li, Z. Murez, D. Kriegman, R. Ramamoorthi and M. Chandraker. Learning to See throughTurbulent Water. IEEE Winter Conference on Applications of Computer Vision, WACV 2018.

14. G. Chen, W. Choi, X. Yu, T. Han and M. Chandraker. Learning Efficient Object DetectionModels with Knowledge Distillation. Neural Information Processing Systems, NIPS 2017.

15. K. Sohn, S. Liu, G. Zhong, X. Yu, M.-H. Yang and M. Chandraker. Unsupervised DomainAdaptation for Face Recognition in Unlabeled Videos. IEEE International Conference on ComputerVision, ICCV 2017.

16. X. Yin, X. Yu, K. Sohn, X. Liu and M. Chandraker. Towards Large-Pose Face Frontalization.IEEE International Conference on Computer Vision, ICCV 2017.

17. X. Peng, X. Yu, K. Sohn, D. Metaxas and M. Chandraker. Feature Reconstruction-Based Dis-entanglement for Pose-Invariant Face Recognition. IEEE International Conference on ComputerVision, ICCV 2017.

18. J. Gwak, C. Choy, A. Garg, M. Chandraker and S. Savarese. Weakly Supervised 3D Recon-struction with Adversarial Constraint. International Conference on 3D Vision, 3DV 2017.

19. Z. Li, Z. Xu, R. Ramamoorthi and M. Chandraker. Robust Energy Minimization for BRDF-Invariant Shape from Light Fields. IEEE Conference on Computer Vision and Pattern Recognition,CVPR 2017.

20. C. Li, Z. Zia, Q.-H. Tran, X. Yu, G. Hager and M. Chandraker. Deep Supervision with ShapeConcepts for Occlusion-Aware 3D Object Parsing. IEEE Conference on Computer Vision andPattern Recognition, CVPR 2017.

21. N. Lee, W. Choi, P. Vernaza, C. Choy, P. Torr and M. Chandraker. DESIRE: Distant FuturePrediction in Dynamic Scenes with Multiple Interacting Agents. IEEE Conference on ComputerVision and Pattern Recognition, CVPR 2017.

22. S. Schulter, P. Vernaza, W. Choi and M. Chandraker. Deep Network Flow for Multi-ObjectTracking. IEEE Conference on Computer Vision and Pattern Recognition, CVPR 2017.

23. L. Zheng, H. Zhang, S. Sun, M. Chandraker, Y. Yang and Q. Tian. Person Re-Identificationin the Wild. IEEE Conference on Computer Vision and Pattern Recognition, CVPR 2017.

‡‡ICCV, CVPR and ECCV are premier conferences in computer vision. Overall accepance rates are about 20% andoral presentations have an acceptance rate of about 3–5%.

24. C. Choy, J. Gwak, S. Savarese and M. Chandraker. Universal Correspondence Network. NeuralInformation Processing Systems, NIPS 2016. [oral presentation]

25. X. Yu, F. Zhou and M. Chandraker. Deep Deformation Network for Object Landmark Local-ization. European Conference on Computer Vision, ECCV 2016. [to appear]

26. T.-C. Wang, E. Hiroaki, J. Zhu, M. Chandraker, A. Efros and R. Ramamoorthi. A 4DLight-Field Dataset and CNN Architectures for Material Recognition. European Conference onComputer Vision, ECCV 2016. [to appear]

27. A. Kanazawa, D. Jacobs and M. Chandraker. WarpNet: Weakly Supervised Matching forSingle-View Reconstruction. IEEE Conference on Computer Vision and Pattern Recognition,CVPR 2016.

28. T.-C. Wang, M. Chandraker, A. Efros and R. Ramamoorthi. SVBRDF-Invariant Shape andReflectance Recovery from Light Fields. IEEE Conference on Computer Vision and PatternRecognition, CVPR 2016. [oral presentation]

29. V. Dhiman, Q.-H. Tran, J. Corso and M. Chandraker. A Continuous Occlusion Model for RoadScene Understanding. IEEE Conference on Computer Vision and Pattern Recognition, CVPR2016.

30. C.-Y. Chen, W. Choi and M. Chandraker. Atomic Scenes for Scalable Traffic Scene Recognition.IEEE Winter Conference on Applications of Computer Vision, WACV 2016.

31. S. Song and M. Chandraker. High Accuracy 3D Object Localization for Autonomous DrivingUsing SFM and Detection Cues. IEEE Conference on Computer Vision and Pattern Recognition,CVPR 2015. [oral presentation]

32. M. Chandraker. On Joint Shape and Material Recovery from Motion Cues. European Confer-ence on Computer Vision, ECCV 2014.

33. M. Chandraker. What Camera Motion Reveals About Shape with Unknown BRDF. IEEEConference on Computer Vision and Pattern Recognition, CVPR 2014. [oral presentation, 4%accepted] [Best Paper Award] ∗∗

34. S. Song and M. Chandraker. Robust Scale Estimation in Real-Time Monocular SFM forAutonomous Driving. IEEE Conference on Computer Vision and Pattern Recognition, CVPR2014.

35. M. Chandraker, D. Reddy, Y. Wang and R. Ramamoorthi. What Motion Reveals AboutShape with Unknown BRDF and Lighting. IEEE Conference on Computer Vision and PatternRecognition, CVPR 2013. [oral presentation, 4% accepted]

36. Y. Bao, M. Chandraker, Y. Lin and S. Savarese. Dense Object Reconstruction with SemanticPriors. IEEE Conference on Computer Vision and Pattern Recognition, CVPR 2013. [oralpresentation, 4% accepted]

37. S. Song and M. Chandraker. Parallel, Real-Time Monocular Visual Odometry. IEEE Confer-ence on Robotics and Automation, ICRA 2013.

38. M. Chandraker and R. Ramamoorthi. What an Image Reveals About Material Reflectance.IEEE International Conference on Computer Vision, ICCV 2011. [oral presentation, 3%accepted]

39. M. Chandraker, J. Bai and R. Ramamoorthi. A Theory of Differential Photometric Stereofor General Isotropic BRDFs. IEEE Conference on Computer Vision and Pattern Recognition,CVPR 2011. [oral presentation, 3.5% accepted]

40. J. Bai, M. Chandraker, T.-T. Ng and R. Ramamoorthi. A Dual Theory of Inverse and ForwardLight Transport. European Conference on Computer Vision, ECCV 2010.

41. M. Chandraker, J. Lim and D.J. Kriegman. Moving in Stereo: Efficient Structure and MotionUsing Lines. IEEE International Conference on Computer Vision, ICCV 2009.

∗∗IEEE CVPR is the highest rated publication venue for computer vision and seventh-highest across all engineeringand computer sciences, according to Google Scholar metrics.

42. M. Chandraker and D.J. Kriegman. Globally Optimal Bilinear Programming for ComputerVision Applications. IEEE Conference on Computer Vision and Pattern Recognition, CVPR 2008.[oral presentation, 4% accepted]

43. M. Chandraker, S. Agarwal, D.J. Kriegman and S. Belongie. Globally Optimal Affine andMetric Upgrades in Stratified Autocalibration. IEEE International Conference on Computer Vision,ICCV 2007. [oral presentation, 4% accepted] [Marr Prize Honorable Mention for BestPaper]∗∗

44. A. Agarwal, S. Izadi, M. Chandraker and A. Blake. High Precision Multi-touch Sensing onSurfaces using Overhead Cameras. IEEE Tabletop 2007.

45. M. Chandraker, S. Agarwal and D.J. Kriegman. ShadowCuts: Photometric Stereo with Shadows.IEEE Conference on Computer Vision and Pattern Recognition, CVPR 2007.

46. M. Chandraker, S. Agarwal, F. Kahl, D. Nister and D.J. Kriegman. Autocalibration via Rank-Constrained Estimation of the Absolute Quadric. IEEE Conference on Computer Vision andPattern Recognition, CVPR 2007.

47. S. Agarwal, M. Chandraker, F. Kahl, D.J. Kriegman and S. Belongie. Practical GlobalOptimization for Multiview Geometry. European Conference on Computer Vision, ECCV 2006.[oral presentation, 5% accepted]

48. M. Chandraker, F. Kahl and D.J. Kriegman. Reflections on the Generalized Bas-ReliefAmbiguity. IEEE Conference on Computer Vision and Pattern Recognition, CVPR 2005. [oralpresentation, 5% accepted]

49. M. Chandraker, C. Stock and A. Pinz. Real-Time Camera Pose in a Room. InternationalConference on Computer Vision Systems, ICVS 2003.

50. C. Stock, U. Muhlmann, M. Chandraker and A. Pinz. Subpixel Corner Detection for TrackingApplications using CMOS Camera Technology. Austrian Association of Pattern Recognition,AAPR 2002.

Refereed Book Chapters

51. M. Chandraker. The Bas-Relief Ambiguity. Computer Vision: A Reference Guide (edited byK. Ikeuchi), pages 43–46.

SELECTED PROFESSIONAL SERVICESArea Chair, AAAI Conference on Artificial Intelligence, AAAI 2019.Workshop Organizer, Autononomous Navigation in Unconstrained Environments, ECCV 2018. AreaChair, IEEE International Conference on Computer Vision, ICCV 2017.Tutorials Chair, International Conference on 3D Vision, 3DV 2016.Area Chair, IEEE Conference on Computer Vision and Pattern Recognition, CVPR 2016.Area Chair, Indian Conf. on Computer Vision, Graphics and Image Processing, 2014, 2016, 2018.

SELECTED RECENT TALKS

Learning Geometry and Semantics for 3D Road Scene Understanding. Keynote, Workshop onApplications of Large-Scale Visual Odometry, CVPR 2018.

Self-Driving: Past, Present and Future. IIIT Workshop on Autonomous Driving in India.

Physical Models for Data-Driven Shape and Material. Keynote, Workshop on Physics-Based DeepLearning, ICCV 2017.

Physically-Motivated CNNs for Material Estimation. Keynote, Workshop on Data-Driven BRDFRepresentations, ICCV 2017.

Insights from Shape and Material for Learning to See in 3D. Keynote, Workshop on Learning toSee from 3D Data, ICCV 2017.

∗∗The Marr Prize is one of the top honors in computer vision, awarded once in two years to the best paper at ICCV.

1

Pavel Pevzner (biographical sketch)

Ronald R. Taylor Chair Professor of Computer Science and Engineering

University of California, San Diego

9500 Gilman Drive, La Jolla, CA 92093-0404

1-310-497-6941

ppevzner@ucsd.edu

(a) Professional Preparation

Moscow Technological Transport Institute, Russia, Applied Mathematics, MS, 1979

Moscow Institute of Physics and Technology, Russia, Mathematics & Physics, Ph.D., 1988

University of Southern California, Los Angeles, Computational Biology, postdoc, 1992

(b) Appointments

2000-present Ronald R. Taylor Chair Professor of Computer Science, UCSD

1995-2000 Professor, Dept. of Mathematics, Computer Science, and Molecular Biology, USC

1992-1995 Assoc. Professor, Dept. of Computer Science, The Pennsylvania State University

1985-1990 Scientist, Lab. of Math. Methods, Natl. Center for Biotechnology, Moscow, Russia

1979-1985 Junior Scientist, Flows in Networks Lab, Natl. Transportation Inst., Moscow, Russia

(c) Products

(i) five most closely related to proposed project

Y. Lin, J. Yuan, M. Kolmogorov, M. W. Shen, M. Chaisson, P. A. Pevzner. Assembly of Long

Error-Prone Reads Using de Bruijn Graphs. Proceedings of the National Academy of Sciences.

(2016) 113:E8396-E8405

D. Antipov, A. Korobeynikov, J.S. McLean, P.A. Pevzner. hybridSPAdes: an algorithm for

hybrid assembly of short and long reads. (2016) Bioinformatics, 32, 1009-1015

Bankevich and P.A. Pevzner. TruSPAdes: barcode assembly of TruSeq synthetic long reads.

(2016) Nature Methods, 13, 248-250

P.A. Pevzner, H. Tang. G. Tesler (2004) De novo repeat classification and fragment assembly.

Genome Res. 14, 1786-1796

P.A. Pevzner, H. Tang, M.S. Waterman (2001) An Eulerian path approach to DNA fragment

assembly Proceedings of the National Academy of Sciences, 98, 9748-9753

(ii) five additional significant products

M.J. Chaisson, P.A. Pevzner (2008) Short read fragment assembly of bacterial genomes. Genome

Res. 18, 324-30.

Chitsaz H, Yee-Greenbaum J.L, Tesler G, Lombardo M.J, Dupont C.L, Badger J.H, Novotny, M,

2

Rusch D.B, Fraser L.J, Gormley N.A, Schulz-Triegla_ O, Smith G.P, Evers D.J, Pevzner, P.A,

Lasken R.S. (2011) Efficient de novo assembly of single-cell bacterial genomes from short-read

data sets. Nature Biotechnology, 29, 915-21

Bankevich A, Nurk S, Antipov D, Gurevich AA, Dvorkin M, Kulikov AS, Lesin VM, Nikolenko

SI, Pham S, Prjibelski AD, Pyshkin AV, Sirotkin AV, Vyahhi N, Tesler G, Alekseyev MA,

Pevzner PA (2012) SPAdes: a new genome assembly algorithms and its applications to single

cell sequencing. J. Computational Biology, 19, 455-77

Prjibelski AD, Vasilinetc I, Bankevich A, Gurevich A, Krivosheeva T, Nurk S, Pham S,

Korobeynikov A, Lapidus A, Pevzner PA. (2014) ExSPAnder: a universal repeat resolver for

DNA fragment assembly. Bioinformatics, 30, 293-301

S. Nurk, Meleshko D, Korobeynikov A, Pevzner PA. metaSPAdes: a new versatile metagenomic

assembler. Genome Res. (2017) 27(5):824-834

(d) Synergistic Activities

Public software: I and my lab members have released many open-source tools that cover genome

assembly, computational mass spectrometry, immunogenomics, and analysis of genome

rearrangements. Our popular SPAdes assembler is now the most cited short read assembler in the

world (over 3300 citations since its release in 2012). The SPAdes family of tools also includes

dipSPAdes (2015), truSPAdes (2016), hybridSPAdes (2016), plasmidSPAdes (2016), and

metaSPAdes (2017). In 2018, we also released the long read assembler Flye.

Textbooks

o Computational Molecular Biology, MIT Press (2000)

o Introduction to Bioinformatics Algorithms” MIT Press (2004)

o Bioinformatics for Biologists” Cambridge University Press (2011)

o Bioinformatics Algorithms: An Active Learning Approach” Active Learning Publishers

(2014)

Massive Open Online Courses (MOOCS)

o Bioinformatics Specialization on Coursera, a series of seven MOOCs (over 300,000

enrollments since 2014). https://www.coursera.org/specializations/bioinformatics

o Algorithms Specialization on Coursera, a series of six MOOCs (over 120,000 enrollments

since February 2016). https://www.coursera.org/specializations/data-structures-

algorithms

o Algorithms and Data Structures MicroMaster Program at edX, a series of eight MOOCs

(launched in January 2018). https://www.edx.org/micromasters/ucsandiegox-algorithms-

and-data-structures

o Introduction to Genomic Data Science MOOC at edX (launched in December 2017)

https://www.edx.org/course/introduction-genomic-data-science-uc-san-diegox-cse181-1x

o Analyze Your Genome! MOOC at EdX (launched in January 2018)

https://www.edx.org/course/analyze-genome-uc-san-diegox-binf180-1

Rosalind web portal for learning bioinformatics through programming (over 50,000 users, used

by over 100 professors annually) http://rosalind.info/problems/locations/

3

Peter Gerstoft

Scripps Institution of Oceanography

University of California, San Diego, La Jolla, CA 92093-0238

noiselab.ucsd.edu, Email: gerstoft@ucsd.edu, Tel: (858) 534-7768

A. Professional Preparation:

1983 MSc., Engineering, Technical University of Denmark, Kgs. Lyngby, Denmark

1984 MSc., Engineering, University of Western Ontario, London, ON, Canada

1986 PhD., Engineering, Technical University of Denmark, Kgs. Lyngby, Denmark

1988 BBA (HD) Finance, Copenhagen School of Economics, Frederiksberg, Denmark

B. Appointments:

2013-present Adj Professor, Electrical and Computer Engineering, University of California, San Diego.

2004-present Data Scientist, Scripps Institution of Oceanography, University of California, San Diego.

1997-2004 Associate Scientist, Marine Physical Laboratory, University of California, San Diego.

1999-2000 Senior Seismic Acoustic Officer, Comprehensive Nuclear-Test-Ban Treaty Organization.

1992-1997 Senior Scientist, NATO SACLANT Undersea Research Centre, La Spezia, Italy.

1989-1990 Visiting Scientist, Dept. of Ocean Engineering, Massachusetts Institute of Technology.

1987-1992 Scientist, Odegaard & Danneskiold-Samsoe Acoustics and Vibrations.

C. Products:

(i) Publications Relevant to the Proposal:

Riahi, Gerstoft (2017), Using Graph Clustering to Locate Sources within a Dense Sensor Array, Signal

Processing 132, March 2017, Pages 110–120

Gerstoft, Bromirski (2016), "Weather bomb" induced seismic signals. Science 26 Aug 2016, vol 353,

Issue 6302, pp. 869-870

Gerstoft, Xenaki, Mecklenbräuker, (2105) Multiple and single snapshot compressive beamforming, J

Acoust. Soc. Am., 138, 2003-2015

Niu, Ozanich, Gerstoft (2017), Ship localization in Santa Barbara Channel using machine learning

classifiers, J Acoust. Soc. Am. 142, EL455-460

Kong, Trugman, Ross, Bianco, Meade, Gerstoft Machine learning in Seismology —Turning data into

insights, Seismological Research Letters,

Gerstoft, Shearer, Harmon, Zhang (2008), Global P, PP, and PKP wave microseisms observed from

distant storms, Geophys. Res. Lett., 35, L23307.

Gerstoft, Fehler, Sabra (2008), When Katrina hit California, Geophys. Res. Lett., 33, L17308, 2006.

Curtis, Gerstoft, Sato, Snieder, Wapenaar. Seismic interferometry--turning noise into signal. The Leading

Edge, 25, 1082-1092, 2006.

Gerstoft, Sabra, Roux, Kuperman, Fehler, Green’s functions extraction and surface wave tomography

from microseisms in southern California, Geophys, 71, SI23-31, 2006.

C. Publications: 170 reviewed papers with > 7500 citations; H-index 47 (Google Scholar). One paper in

Science and one in PNAS. For more publications see noiselab.ucsd.edu, or Gerstoft in Google Scholar.

D. Synergistic Activities:

(i) Developed widely used computer codes for the scientific community: a. SAGA (1992- present) - General geoacoustic and electromagnetic inversion program:

approximately 100 installations performed in twenty countries. noiselab.ucsd.edu/saga/saga.html

b. CABRILLO (1997-1999) – Seismic exploration finite difference modeling of acoustic, elastic

and poroelastic media. http://noiselab.ucsd.edu/cabrillo/cabrillo.html

c. OSIRIS (1987-1992) - Commercial seismic exploration software for modeling of synthetic

seismograms.

(ii) Organizer and Chairman of:

a. Technical Chair, Acoustical Society of America, San Diego, 2011.

Chair, Acoustical Society of America, San Diego, 2019.

b. Special sessions on

- Inversion using noise sources, Acoustical Society of America Fall, Minneapolis 2005.

- Geoacoustic inversion at the Underwater Acoustic Measurements Conference, Crete 2005

- Geoacoustic inversion at the Underwater Acoustic Measurements Conference, Crete 2007

- Geoacoustic inversion at the Underwater Acoustic Measurements Conference, Greece 2009

- Geoacoustic inversion at the Acoustical Society of America Spring 2008 Paris meeting.

- Signal processing session at the Acoustical Society of America Spring 2009 Meeting, Portland.

- Kalman and particle filters at the Acoustical Society of America Fall 2009 Meeting, San Antonio.

- Bottom characterization and geoacoustic inversion, European Conference on Ocean acoustics, 2010.

- Bottom characterization at the Underwater Acoustic Measurements Conference, Greece 2011.

- Array signal processing at the IEEE Asilomar Conference on Signals, Systems, and Computers, 2012.

- Properties, Trends, and Utilization of Ocean Noise, Acoustical Soc America fall 2013 San Francisco.

- Sonar signal and information processing, Underwater Acoustic Measurements Conference, Greece 2014.

- Sonar signal and information processing, Underwater Acoustic Measurements Conference, Greece, 2015.

- Remote sensing with noise, Information Theory with Applications, San Diego, 2015

- Sparse sampling, Information Theory with Applications, San Diego, 2016

- Compressive sensing in acoustics, Acoustical Soc America fall 2016 Honolulu.

- Graph Signal Processing, Information Technology and Applications, San Diego 2018

(iii) PhD Thesis Advisor: William Jenkins (2017-), Mike Bianco (2015-18), Emma Reeves (2014-), Mark Wagner (2014-) Zhao Chen

(2012-18), Jie Li (2015-17), Angeliki Xenaki (2013-15), Anup Das (2014-15), Wenyuan Fan (2011-14), Erich

Zoechman (2013-14), Daniel Bien Aik Tan (2009-14), Florian Xaver (2008-13), Ravi Menon (2008-13) [Data

Scientist, Slack], Ali Kaiman (2009-12), James Traer (2006-11) [MIT], Tim Ray (2008-10), Laura Brooks

(2006-08) [Prof Adelaide University], Yong Han Goh (2006-07) [DSO in Singapore], Caglar Yardim (2003-

07) [Ohio state U], Chen-fen Huang (2001-05)[Prof, Natl Taiwan Univ]

(iv) Postgraduate Sponsor: Mike Bianco (2018-) Santosh Nannuru (2016-), Kai Gemba (2015-), Haiquang Nui (2015-17), Nima Riahi

(2014-16), Anja Diez (2015-16), Ravi Menon (2014), Olivier Carriere (2011-13), James Traer (2011-12),

Huajian Yao (2010-12), Michael Lewis (2009-11), Jian Zhang (2007-11), Caglar Yardim (2007-09), Chen-

Fen Huang (2005-07), Karim Sabra (2003-05), David Battle (2001-03), Kaelig Castor (2002-04)

Memberships: Fellow Acoustical Society of America. Elected member of the International Union of

Radio Science, Commission F. Senior Member IEEE.

Philip Guo

Assistant ProfessorDepartment of Cognitive ScienceUniversity of California, San Diego (UCSD)

Updated: July 27, 2018Email: philip@pgbovine.netwww.pgbovine.net

RESEARCH INTERESTS

Human-computer interaction, online learning at scale, productivity tools for computerprogrammers and data scientists, computing education

ACADEMIC POSITIONS

07/2016 – University of California, San Diego (UCSD), La Jolla, CAAssistant Professor of Cognitive ScienceAffiliate Assistant Professor of Computer Science and EngineeringFaculty Affiliate: Design Lab, Halicioglu Data Science Institute

07/2014 – 06/2016 University of Rochester, Rochester, NYAssistant Professor of Computer Science (2014–2016)Research Assistant Professor of Computer Science (2016–present)

EDUCATION

09/2006 – 06/2012 Stanford University, Stanford, CAPh.D. in Computer ScienceDissertation: Software Tools to Facilitate Research ProgrammingAdvisor: Dawson Engler

06/2005 – 06/2006 Massachusetts Institute of Technology, Cambridge, MAMaster of Engineering in Electrical Engineering and Computer ScienceMaster’s Thesis: A Scalable Mixed-Level Approach to Dynamic Analysis of C andC++ Programs, Advisor: Michael D. Ernst(MIT EECS award for Outstanding Computer Science Master of Engineering Thesis)

09/2001 – 06/2005 Massachusetts Institute of Technology, Cambridge, MABachelor of Science in Electrical Engineering and Computer Science, GPA: 5.0/5.0

AWARDS AND HONORS

04/2018 CHI Honorable Mention Paper Award [C.38]

10/2017 UIST Honorable Mention Paper Award [C.36]

05/2017 CHI Honorable Mention Paper Award [C.31]

08/2015 Google Faculty Research Award

04/2014 CHI Honorable Mention Paper Award [C.18]

06/2012 ICSE Software Engineering In Practice Best Paper Award [C.13]

07/2009 ACM SIGSOFT Distinguished Paper Award [C.6]

04/2009 CHI Honorable Mention Paper Award [C.3]

Philip Guo – Page 1 of 11 – Curriculum Vitae

09/2009 – 06/2011 National Science Foundation (NSF) Graduate Fellowship

09/2006 – 09/2009 National Defense Science and Engineering (NDSEG) Graduate Fellowship

05/2006 MIT Charles and Jennifer Johnson Thesis Award for Outstanding Computer ScienceMaster of Engineering Thesis

PRIOR EMPLOYMENT

07/2015 – 08/2015 Microsoft Research, Redmond, WAVisiting Researcher – Research in Software Engineering (RiSE) group

10/2013 – 06/2014 Massachusetts Institute of Technology, Cambridge, MAPostdoctoral Researcher – CSAIL User Interface Design Group – Host: Rob Miller

06/2013 – 09/2013 edX, Cambridge, MAVisiting Research Scientist – analyzed MOOC data [C.15, C.16, C.17]

07/2012 – 02/2013 Google, Mountain View, CASoftware Engineer – online education group – Google Research

09/2006 – 06/2012 Stanford University, Stanford, CAPh.D. Student – Department of Computer Science

09/2011 – 01/2012 Harvard University, Cambridge, MAVisiting Research Fellow – Computer Systems Group – Host: Margo Seltzer

06/2011 – 09/2011 Google, Mountain View, CASoftware Engineering Intern – refined and deployed CDE [C.9, C.12, M.2, B.2]

06/2009 – 09/2009 Microsoft Research, Redmond, WAResearch Intern – Research in Software Engineering (RiSE) group

06/2007 – 09/2007 Google, Mountain View, CASoftware Engineering Intern – prototyped memory allocators for C and C++ programs

01/2004 – 06/2006 Massachusetts Institute of Technology, Cambridge, MAResearch Assistant – Program Analysis Group – Advisor: Michael D. ErnstUndergraduate and master’s research on tools for analyzing C and C++ programs

09/2003 – 01/2004 Massachusetts Institute of Technology, Cambridge, MAResearch Assistant – Computer Graphics Group – Advisor: Fredo DurandDeveloped an HDR (high dynamic range) image editing tool for photographers

06/2004 – 08/2004,06/2003 – 08/2003

Teradyne, Agoura Hills, CASoftware Engineering Intern – wrote simulators for semiconductor test hardware

09/2002 – 06/2003 Massachusetts Institute of Technology, Cambridge, MAResearch Assistant – Teacher Education Program – Advisor: Eric KlopferDeveloped a suite of 5 educational games for Palm OS devices

06/2002 – 08/2002 Codehost, Culver City, CASoftware Engineering Intern – wrote embedded Linux tablet PC software

Philip Guo – Page 2 of 11 – Curriculum Vitae

FUNDING

National Science Foundation. NRT-IGE: Augmenting, Piloting, and Scaling Com-putational Notebooks to Train New Graduate Researchers in Data-Centric Program-ming. $498,751 (co-PI, 2017–2020. PI: James Hollan, co-PI: Philip Guo, co-PI: ScottKlemmer, co-PI: Bradley Voytek)

National Science Foundation. CRII: CHS: Scaling Up Online Peer Tutoring of Com-puter Programming. $175,000 (sole PI, 2015–2018)

NSF CRII REU (Research Experiences for Undergraduates) supplement. $32,000

Google Faculty Research Award. Enabling Learners to Create Hierarchical Tutorialsfrom How-To Videos on YouTube. $64,295 (sole PI, 2015)

University of Rochester. University Research Award: Enabling Fast and ScalableFeedback on Writing. $50,000 (sole PI, 2015)

Microsoft Research. Online Python Tutor for Office Mix. $61,308 (sole PI, 2014)

PEER-REVIEWED PUBLICATIONS

Note that in many areas within computer science and human-computer interaction,conferences (not journals) are the primary venues for peer-reviewed publications.

ConferencePapers

C.43 Xiong Zhang and Philip J. Guo. Fusion: Opportunistic Web Prototyping withUI Mashups. In Proceedings of UIST 2018: ACM Symposium on User InterfaceSoftware and Technology, Oct 2018.

C.42 Alok Mysore and Philip J. Guo. Porta: Profiling Software Tutorials UsingOperating-System-Wide Activity Tracing. In Proceedings of UIST 2018: ACMSymposium on User Interface Software and Technology, Oct 2018.

C.41 Kyle Thayer, Philip J. Guo, Katharina Reinecke. The Impact of Culture onLearner Behavior in Visual Debuggers. In Proceedings of VL/HCC 2018: IEEESymposium on Visual Languages and Human-Centric Computing, Oct 2018.

C.40 Kandarp Khandwala and Philip J. Guo. Codemotion: Expanding the DesignSpace of Learner Interactions with Computer Programming Tutorial Videos. InProceedings of L@S 2018: ACM Conference on Learning at Scale, June 2018.

C.39 Sean Kross and Philip J. Guo. Students, Systems, and Interactions: Synthe-sizing the First Four Years of Learning@Scale and Charting the Future. InProceedings of L@S 2018: ACM Conference on Learning at Scale, June 2018.

C.38 April Y. Wang, Ryan Mitts, Philip J. Guo, Parmit K. Chilana. Mismatch ofExpectations: How Modern Learning Resources Fail Conversational Program-mers. In Proceedings of CHI 2018: ACM Conference on Human Factors inComputing Systems, April 2018.(Honorable Mention Paper Award)

C.37 Philip J. Guo. Non-Native English Speakers Learning Computer Programming:Barriers, Desires, and Design Opportunities. In Proceedings of CHI 2018: ACMConference on Human Factors in Computing Systems, April 2018.

Philip Guo – Page 3 of 11 – Curriculum Vitae

C.36 Xiong Zhang and Philip J. Guo. DS.js: Turn Any Webpage into an Example-Centric Live Programming Environment for Learning Data Science. In Pro-ceedings of UIST 2017: ACM Symposium on User Interface Software and Tech-nology, Oct 2017.(Honorable Mention Paper Award)

C.35 Hyeonsu Kang and Philip J. Guo. Omnicode: A Novice-Oriented Live Pro-gramming Environment with Always-On Run-Time Value Visualizations. InProceedings of UIST 2017: ACM Symposium on User Interface Software andTechnology, Oct 2017.

C.34 Alok Mysore and Philip J. Guo. Torta: Generating Mixed-Media GUI andCommand-Line App Tutorials Using Operating-System-Wide Activity Tracing.In Proceedings of UIST 2017: ACM Symposium on User Interface Software andTechnology, Oct 2017.

C.33 Ian Drosos, Philip J. Guo, Chris Parnin. HappyFace: Identifying and PredictingFrustrating Obstacles for Learning Programming at Scale. In Proceedings ofVL/HCC 2017: IEEE Symposium on Visual Languages and Human-CentricComputing, Oct 2017.

C.32 Jeremy Warner and Philip J. Guo. Hack.edu: Examining How College HackathonsAre Perceived By Student Attendees and Non-Attendees. In Proceedings ofICER 2017: ACM International Computing Education Research conference,August 2017.

C.31 Philip J. Guo. Older Adults Learning Computer Programming: Motivations,Frustrations, and Design Opportunities. In Proceedings of CHI 2017: ACMConference on Human Factors in Computing Systems, May 2017.(Honorable Mention Paper Award)

C.30 Jeremy Warner and Philip J. Guo. CodePilot: Scaffolding End-to-End Collab-orative Software Development for Novice Programmers. In Proceedings of CHI2017: ACM Conference on Human Factors in Computing Systems, May 2017.

C.29 Denae Ford, Justin Smith, Philip J. Guo, Chris Parnin. Paradise Unplugged:Identifying Barriers for Female Participation on Stack Overflow. In Proceedingsof FSE 2016: ACM SIGSOFT International Symposium on the Foundations ofSoftware Engineering, Nov 2016.

C.28 Parmit K. Chilana, Rishabh Singh, Philip J. Guo. Understanding Conversa-tional Programmers: A Perspective from the Software Industry. In Proceedingsof CHI 2016: ACM Conference on Human Factors in Computing Systems, May2016.

C.27 Philip J. Guo. Codeopticon: Real-Time, One-To-Many Human Tutoring forComputer Programming. In Proceedings of UIST 2015: ACM Symposium onUser Interface Software and Technology, Nov 2015.

C.26 Philip J. Guo, Jeffery White, Renan Zanelatto. Codechella: Multi-User Pro-gram Visualizations for Real-Time Tutoring and Collaborative Learning. InProceedings of VL/HCC 2015: IEEE Symposium on Visual Languages andHuman-Centric Computing, Oct 2015.

C.25 Mitchell Gordon and Philip J. Guo. Codepourri: Creating Visual Coding Tuto-rials Using A Volunteer Crowd Of Learners. In Proceedings of VL/HCC 2015:IEEE Symposium on Visual Languages and Human-Centric Computing, Oct2015.

Philip Guo – Page 4 of 11 – Curriculum Vitae

C.24 Joyce Zhu, Jeremy Warner, Mitchell Gordon, Jeffery White, Renan Zanelatto,Philip J. Guo. Toward a Domain-Specific Visual Discussion Forum for Learn-ing Computer Programming: An Empirical Study of a Popular MOOC Forum.In Proceedings of VL/HCC 2015: IEEE Symposium on Visual Languages andHuman-Centric Computing, Oct 2015.

C.23 Parmit K. Chilana, Celena Alcock, Shruti Dembla, Anson Ho, Ada Hurst, BrettArmstrong, Philip J. Guo. Perceptions of Non-CS Majors in Intro Program-ming: The Rise of the Conversational Programmer. In Proceedings of VL/HCC2015: IEEE Symposium on Visual Languages and Human-Centric Computing,Oct 2015.

C.22 Jeremy Warner, John Doorenbos, Bradley N. Miller, Philip J. Guo. How HighSchool, College, and Online Students Differentially Engage with an InteractiveDigital Textbook. Short paper in Proceedings of EDM 2015: InternationalConference on Educational Data Mining, June 2015.

C.21 Carrie J. Cai, Philip J. Guo, James Glass, Robert C. Miller. Wait-Learning:Leveraging Wait Time for Second Language Education. In Proceedings of CHI2015: ACM Conference on Human Factors in Computing Systems, April 2015.

C.20 Juho Kim, Philip J. Guo, Carrie J. Cai, Shang-Wen (Daniel) Li, KrzysztofZ. Gajos, Robert C. Miller. Data-Driven Interaction Techniques for Improv-ing Navigation of Educational Videos. In Proceedings of UIST 2014: ACMSymposium on User Interface Software and Technology, October 2014.

C.19 Jeremy Scott, Philip J. Guo, Randall Davis. A Direct Manipulation Languagefor Explaining Algorithms. Short paper in Proceedings of VL/HCC 2014: IEEESymposium on Visual Languages and Human-Centric Computing, Jul 2014.

C.18 Juho Kim, Phu Nguyen, Sarah Weir, Philip J. Guo, Robert C. Miller, KrzysztofZ. Gajos. Crowdsourcing Step-by-Step Information Extraction to EnhanceExisting How-to Videos. In Proceedings of CHI 2014: ACM Conference onHuman Factors in Computing Systems, April 2014.(Honorable Mention Paper Award)

C.17 Philip J. Guo and Katharina Reinecke. Demographic Differences in How Stu-dents Navigate Through MOOCs. In Proceedings of L@S 2014: ACM Confer-ence on Learning at Scale, March 2014.

C.16 Philip J. Guo, Juho Kim, Rob Rubin. How Video Production Affects StudentEngagement: An Empirical Study of MOOC Videos. In Proceedings of L@S2014: ACM Conference on Learning at Scale, March 2014.

C.15 Juho Kim, Philip J. Guo, Daniel T. Seaton, Piotr Mitros, Krzysztof Z. Gajos,Robert C. Miller. Understanding In-Video Dropouts and Interaction Peaksin Online Lecture Videos. In Proceedings of L@S 2014: ACM Conference onLearning at Scale, March 2014.

C.14 Philip J. Guo. Online Python Tutor: Embeddable Web-Based Program Visu-alization for CS Education. In Proceedings of SIGCSE 2013: ACM TechnicalSymposium on Computer Science Education, March 2013.

C.13 Thomas Zimmermann, Nachiappan Nagappan, Philip J. Guo, Brendan Murphy.Characterizing and Predicting Which Bugs Get Reopened. In Proceedings ofICSE 2012: ACM/IEEE International Conference on Software Engineering,Software Engineering In Practice track, June 2012.(Best Paper Award)

Philip Guo – Page 5 of 11 – Curriculum Vitae

C.12 Philip J. Guo. CDE: Run Any Linux Application On-Demand Without In-stallation. In Proceedings of LISA 2011: USENIX Large Installation SystemAdministration Conference, December 2011.

C.11 Philip J. Guo, Sean Kandel, Joseph M. Hellerstein, Jeffrey Heer. ProactiveWrangling: Mixed-Initiative End-User Programming of Data TransformationScripts. In Proceedings of UIST 2011: ACM Symposium on User InterfaceSoftware and Technology, October 2011.

C.10 Philip J. Guo and Dawson Engler. Using Automatic Persistent Memoizationto Facilitate Data Analysis Scripting. In Proceedings of ISSTA 2011: ACMInternational Symposium on Software Testing and Analysis, July 2011.

C.9 Philip J. Guo and Dawson Engler. CDE: Using System Call Interposition toAutomatically Create Portable Software Packages. Short paper in Proceedingsof USENIX 2011: USENIX Annual Technical Conference, June 2011.

C.8 Philip J. Guo, Thomas Zimmermann, Nachiappan Nagappan, Brendan Murphy.“Not My Bug!” and Other Reasons for Software Bug Report Reassignments.In Proceedings of CSCW 2011: ACM Conference on Computer Supported Co-operative Work, March 2011.

C.7 Philip J. Guo, Thomas Zimmermann, Nachiappan Nagappan, Brendan Murphy.Characterizing and Predicting Which Bugs Get Fixed: An Empirical Study ofMicrosoft Windows. In Proceedings of ICSE 2010: ACM/IEEE InternationalConference on Software Engineering, May 2010.

C.6 Adam Kiezun, Vijay Ganesh, Philip J. Guo, Pieter Hooimeijer, Michael D.Ernst. HAMPI: A Solver for String Constraints. In Proceedings of ISSTA:ACM International Symposium on Software Testing and Analysis, July 2009.(ACM SIGSOFT Distinguished Paper Award)

C.5 Philip J. Guo and Dawson Engler. Linux Kernel Developer Responses to StaticAnalysis Bug Reports. Short paper in Proceedings of USENIX 2009: USENIXAnnual Technical Conference, June 2009.

C.4 Adam Kiezun, Philip J. Guo, Karthick Jayaraman, Michael D. Ernst. Auto-matic Creation of SQL Injection and Cross-site Scripting Attacks. In Proceed-ings of ICSE 2009: ACM/IEEE International Conference on Software Engi-neering, May 2009.

C.3 Joel Brandt, Philip J. Guo, Joel Lewenstein, Mira Dontcheva, Scott R. Klem-mer. Two Studies of Opportunistic Programming: Interleaving Web Foraging,Learning, and Writing Code. In Proceedings of CHI 2009: ACM Conferenceon Human Factors in Computing Systems, April 2009.(Honorable Mention Paper Award)

C.2 Philip J. Guo, Jeff H. Perkins, Stephen McCamant, Michael D. Ernst. Dy-namic Inference of Abstract Types. In Proceedings of ISSTA 2006: ACMInternational Symposium on Software Testing and Analysis, July 2006.

C.1 Brian Demsky, Michael D. Ernst, Philip J. Guo, Stephen McCamant, Jeff H.Perkins, Martin Rinard. Automatic Inference and Enforcement of Data Struc-ture Consistency Specifications. In Proceedings of ISSTA 2006: ACM Interna-tional Symposium on Software Testing and Analysis, July 2006.

Philip Guo – Page 6 of 11 – Curriculum Vitae

JournalArticles

J.3 Elena L. Glassman, Jeremy Scott, Rishabh Singh, Philip J. Guo, Robert C.Miller. OverCode: Visualizing Variation in Student Solutions to ProgrammingProblems at Scale. In TOCHI: ACM Transactions on Computer-Human Inter-action, 2015.

J.2 Adam Kiezun, Vijay Ganesh, Shay Artzi, Philip J. Guo, Pieter Hooimeijer,Michael D. Ernst. Hampi: A Solver for Word Equations over Strings, RegularExpressions and Context-free Grammars. In TOSEM: ACM Transactions onSoftware Engineering Methodology, 2012.

J.1 Michael D. Ernst, Jeff H. Perkins, Philip J. Guo, Stephen McCamant, CarlosPacheco, Matthew S. Tschantz, Chen Xiao. The Daikon system for dynamicdetection of likely invariants. In Science of Computer Programming, 2007.

WorkshopPapers

W.4 Philip J. Guo and Margo Seltzer. Burrito: Wrapping Your Lab Notebook inComputational Infrastructure. In Proceedings of TaPP 2012: USENIX Work-shop on the Theory and Practice of Provenance, June 2012.

W.3 Philip J. Guo. Sloppy Python: Using Dynamic Analysis to Automatically AddError Tolerance to Ad-Hoc Data Processing Scripts. In Proceedings of WODA2011: ACM International Workshop on Dynamic Analysis, July 2011.

W.2 Philip J. Guo and Dawson Engler. Towards Practical Incremental Recompu-tation for Scientists: An Implementation for the Python Language. In Pro-ceedings of TaPP 2010: USENIX Workshop on the Theory and Practice ofProvenance, February 2010.

W.1 Joel Brandt, Philip J. Guo, Joel Lewenstein, Scott R. Klemmer. OpportunisticProgramming: How Rapid Ideation and Prototyping Occur in Practice. InWEUSE 2008: ACM Workshop on End-User Software Engineering, May 2008.

Posters andWorks-in-Progress

P.3 Elena L. Glassman, Jeremy Scott, Rishabh Singh, Philip J. Guo, Robert C.Miller. OverCode: Visualizing Variation in Student Solutions to ProgrammingProblems at Scale. Poster in Proceedings of UIST 2014: ACM Symposium onUser Interface Software and Technology, October 2014.

P.2 Carrie J. Cai, Philip J. Guo, James Glass, Robert C. Miller. Wait-Learning:Leveraging Conversational Dead Time for Second Language Education. InProceedings of CHI 2014: ACM Conference on Human Factors in ComputingSystems, April 2014.

P.1 Anvisha Pai, Philip J. Guo, Robert C. Miller. Modeling Programming Knowl-edge for Mentoring at Scale. In Proceedings of L@S 2014: ACM Conferenceon Learning at Scale, March 2014.

INVITED PUBLICATIONS

MagazineArticles

M.12 Philip J. Guo. Building Tools to Help Students Learn to Program. In Commu-nications of the ACM, Vol. 60, No. 12, Dec 2017.

M.11 Philip J. Guo. How adults ages 60+ are learning to code. In Communicationsof the ACM, Vol. 60, No. 8, Aug 2017.

M.10 Philip J. Guo. Learning Programming at Scale. In O’Reilly Radar, Aug 2015.

M.9 Philip J. Guo. Refining Students’ Coding and Reviewing Skills. In Communi-cations of the ACM, Vol. 57, No. 9, Sep 2014.

Philip Guo – Page 7 of 11 – Curriculum Vitae

M.8 Philip J. Guo. The Difficulty of Teaching Programming Languages, and theBenefits of Hands-on Learning. In Communications of the ACM, Vol. 57, No.7, Jul 2014. (appeared alongside an unrelated article by Mark Guzdial)

M.7 Philip J. Guo. Clarifying Human-Computer Interaction. In Communicationsof the ACM, Vol. 57, No. 2, Feb 2014.

M.6 Philip J. Guo. Silent Technical Privilege. In Slate, Jan 2014.

M.5 Philip J. Guo. Helping scientists, engineers to work up to 100 times faster. InCommunications of the ACM, Vol. 56, No. 10, Oct 2013.

M.4 Philip J. Guo. Teaching Programming the Way It Works Outside the Class-room. In Communications of the ACM, Vol. 56, No. 8, Aug 2013.

M.3 Philip J. Guo. Lessons from the Grind: How unglamorous grunt work can leadto creative innovation. In MIT Technology Review, Jan 2013.

M.2 Philip J. Guo. CDE: A Tool For Creating Portable Experimental SoftwarePackages. In Computing in Science and Engineering: Special Issue on Softwarefor Reproducible Computational Science, Jul/Aug 2012.

M.1 Joel Brandt, Philip J. Guo, Joel Lewenstein, Mira Dontcheva, Scott R. Klem-mer. Opportunistic Programming: Writing Code to Prototype, Ideate, andDiscover. In IEEE Software: Special Issue on End-User Software Engineering,Sep/Oct 2009.

BookChapters

B.3 Philip J. Guo. Parse that data! Practical Tips for preparing your raw datafor analysis. Book chapter in Perspectives on Data Science for Software En-gineering, T. Menzies, L. Williams, T. Zimmermann, eds. Morgan Kaufmann,2016.

B.2 Philip J. Guo. CDE: Automatically Package and Reproduce ComputationalExperiments. Book chapter in Implementing Reproducible Research, V. Stod-den, F. Leisch, R. Peng, eds. Taylor & Francis Group, 2013.

B.1 Joel Brandt, Philip J. Guo, Joel Lewenstein, Mira Dontcheva, Scott R. Klem-mer. How the Web Helps People Turn Ideas Into Code. Book chapter inNo Code Required: Giving Users Tools to Transform the Web, A. Cypher, M.Dontcheva, T. Lau, J. Nichols, eds. Morgan Kaufmann, 2010.

InvitedPapers

IP.1 Quanzeng You, Jianbo Yuan, Jiaqi Wang, Philip J. Guo, Jiebo Luo. Snapn’ Shop: Visual Search-Based Mobile Shopping Made a Breeze by Machineand Crowd Intelligence. In Proceedings of ICSC 2015: IEEE InternationalConference on Semantic Computing, Feb 2015.

INVITED TALKS

• Learning Programming at Scale. Coursera, Berkeley Institute for Data Science,South Park Commons, Google, May 2018; Caltech CMS Department, April 2018.

• The Design Space of Tools for Learning Programming at Scale.UCSD Design at Large Seminar, October 2016.

• Interactive Systems for Learning Programming at Scale (faculty candidate jobtalk).Northwestern University EECS and School of Education and Social Policy, UCSDCognitive Science, CU Boulder CS, Yale University CS, UCLA CS, UC BerkeleySchool of Information, UCSD Computer Science & Engineering, Jan–Apr 2016.

Philip Guo – Page 8 of 11 – Curriculum Vitae

• Interactive Systems for Learning Programming at Scale. Stanford Human-ComputerInteraction Seminar, Stanford, CA, Feb 2016; Recurse Center (nee Hacker School),New York, NY, Jan 2016; University of Maryland, College Park, MD, Dec 2015.

• Invited panelist on tools for personalized education, CCC visioning workshop onComputer-Aided Personalized Education, Washington, D.C., Nov 2015.

• Learning Programming at Scale. University of Rochester Laboratory for LaserEnergetics, Rochester, NY, Sep 2015; Microsoft Research, Redmond, WA, Aug2015; University of Washington DUB seminar, Seattle, WA, July 2015.

• Online Python Tutor: A 5-Year Retrospective.Union College, Rensselaer Polytechnic Institute (RPI), New York, Oct 2014.

• How to effectively ask for help as a junior employee.MIT 6.UAT guest lecture, Cambridge, MA, Nov 2013.

• Hacking the Ph.D.: Three Serendipitous Projects.Hacker School, New York, NY, Nov 2013.

• Why Pursue A Ph.D.? Three Practical Reasons. Amherst College, UMass Amherst,Brown, MIT, Harvard, Tufts. Oct–Nov 2013.

• Challenges in Teaching Python Programming: Vocabulary, Meaning, and Idioms.MIT Lincoln Laboratory, Lexington, MA, Oct 2013.

• Twenty Lessons From The Ph.D. Grind.Keynote at the MIT CSAIL Student Workshop, Oct 2013.

• Software Tools for Research Programming. Boston University, MA, Sep 2013.

• Programming On Demand: Wrangling, Iterating, and Opportunistic Learning.(faculty candidate job talk, all in CS or EECS departments)University of Utah, North Carolina State University, Dartmouth College, Univer-sity of San Francisco, Oregon State University, Northeastern University, Univer-sity of Rochester, Washington University in St. Louis, Feb–Mar 2013.

• Online Python Tutor: Web-Based Program Visualization for CS Education. SonomaState University – Computer Science Colloquium, Rohnert Park, CA, Nov 2012;Hacker School, New York, NY, Oct 2012.

• The Ph.D. Grind: Candid Discussions About Ph.D. Life. UC Riverside – ComputerScience Colloquium, Riverside, CA, Oct 2012; Google Tech Talk, Mountain View,CA, August 2012.

• CDE: automatically creating reproducible experimental software packages. Repro-ducible Research: Tools and Strategies for Scientific Computing interdisciplinarymeeting, Vancouver, Canada, July 2011; NASA JPL, Pasadena, CA, May 2011.

• CDE: Using System Call Interposition to Automatically Create Portable SoftwarePackages. Google Tech Talk, Mountain View, CA, Feb 2011.

• The potentials and challenges of implementing automatic test generation usingcombined concrete and symbolic execution. Fujitsu, Sunnyvale, CA, Oct 2009.

• Automatic Creation of SQL Injection and Cross-site Scripting Attacks. SamsungR&D Center, San Jose, CA, May 2009.

Philip Guo – Page 9 of 11 – Curriculum Vitae

SERVICE

ProgramCommittee Member

• CHI 2019 (ACM Conference on Human Factors in Computing Systems)• UIST 2018 (ACM Symposium on User Interface Software and Technology)• VL/HCC 2018 (Symposium on Visual Languages & Human-Centric Computing)• L@S 2018 (ACM Conference on Learning at Scale)• EDM 2018 (International Conference on Educational Data Mining)• LIVE 2018 (Workshop on Live Programming, located at SPLASH 2018)• ICER 2017 review committee (Int’l Computing Education Research Conference)• LIVE 2017 (Workshop on Live Programming, located at SPLASH 2017)• EDM 2017 (International Conference on Educational Data Mining)• VISSOFT 2016 (IEEE Working Conference on Software Visualization)• EDM 2016 (International Conference on Educational Data Mining)• L@S 2016 (ACM Conference on Learning at Scale)• L@S 2015 (ACM Conference on Learning at Scale)• PLOOC 2015 (Workshop on Programming Languages Technology for MOOCs)• CHESE 2015 (Int’l Code Hunt Workshop on Educational Software Engineering)• SPLASH-E 2015 (Systems, Programming, Languages and Applications: Software

for Humanity – Education Symposium)• PLATEAU 2012 (Workshop on Evaluation and Usability of Programming Lan-

guages and Tools)• TaPP 2012 (Workshop on the Theory and Practice of Provenance)

External PaperReviewer

CHI (2014–2018), UIST (2013–2017), CSCW (2014–2018), TOCE (2017–2018), Com-puter Science Education (2017), JSME (2017), JVLC (2016), VLSS (2016), IEEE Soft-ware (2016), TOCHI (2015–2016,2018), IUI (2015), MobileHCI (2015–2016), UbiComp(2015), JAIED (2015,2017), TSE (2014–2015,2017) PLDI (2013), EuroSys (2012),POPL (2011), ECOOP (2006, 2009)

Grant Reviewer Sloan Foundation (2012), NSF panel (2016, 2017)(details omitted for confidentiality)

Artifact Eval.Committee

ESEC/FSE 2011 (Symposium on the Foundations of Software Engineering)

TEACHING

Instructor • UCSD COGS 120/CSE 170: Human-Computer Interaction Design(Fall 2016, Fall 2017)

• UCSD COGS 124: Human-Computer Interaction Technical Systems Research(Fall 2017)

• UCSD CSE 219/COGS 229/DSGN 119: Design at Large seminar series(Fall 2017)

• UCSD COGS 121: Human-Computer Interaction Programming Studio(Spring 2017, Spring 2018)

• UCSD COGS 231: Grad Seminar on Human-Centered Programming(Spring 2017, Spring 2018)

• University of Rochester CSC 210: Principles of Web Application Development(Fall 2014, Fall 2015)

• University of Rochester CSC 253: Dynamic Languages & Software Development(Fall 2014)

• MIT 6.813 – User Interface Design & Implementation, co-taught with Rob Miller,Daniel Jackson, and David Karger (Spring 2014)

TeachingAssistant

• Stanford CS343 – Advanced Topics in Compilers (Spring 2012)

Philip Guo – Page 10 of 11 – Curriculum Vitae

• Stanford CS242 – Programming Languages (Autumn 2009)• Stanford CS243 – Advanced Compiling Techniques (Winter 2008)• MIT 6.170 – Laboratory in Software Engineering (Spring 2006)

UndergraduateLaboratoryAssistant

• MIT 6.170 – Laboratory in Software Engineering (Fall 2004)• MIT 6.111 – Introductory Digital Systems Laboratory (Fall 2004)• MIT 6.001 – Structure and Interpretation of Computer Programs (Spring 2002)

CURRENT AND FORMER RESEARCH STUDENTS SUPERVISED

Ph.D. • Ian Drosos [C.33]• Logan Gittelson• Sean Kross [C.39]• Jaime Montoya• Xiong Zhang [C.36,C.43]

Masters • Davide Berdin (visiting student from Uppsala University, Sweden)• Hsien-che (Charles) Chen – undergrad+masters• Hyeonsu Kang [C.35] – first position: research engineer at MIT, then Ph.D. student

at CMU Human-Computer Interaction Institute• Kandarp Khandwala [C.40]• Alok Mysore [C.34,C.42] – 2017 UCSD CSE masters student research award win-

ner, first position: Yelp• Dan Scarafoni – undergrad+masters, first position: MIT Lincoln Laboratory, then

Ph.D. student in Machine Learning at Georgia Tech• Jeremy Warner [C.22,C.24,C.30,C.32] – undergrad+masters, first position: Ph.D.

student at UC Berkeley EECS Department• Jeffery White [C.24,C.26]• Renan Zanelatto [C.24,C.26]

Undergraduate • Karina Banda• Lenny Brown• Irene Chen – first position: Google• Jennifer (Kate) Godzicki• Mitchell Gordon [C.24,C.25] – 2015 CRA Outstanding Undergraduate Researcher

Award winner, first position: Ph.D. student at Stanford Computer Science Dept.• Dan Hassin• Sara Lickers• Emy Lin – first position: Intel• Douglas Miller – first position: Jump Trading• Anvisha Pai – first position: Dropbox• Annie Zhang• Joyce Zhu [C.24] – 2015 CRA Outstanding Undergraduate Researcher Award hon-

orable mention, first position: Quip

Ph.D. CommitteeMember

Erin Brady, Anna Loparev, Phyo Thiha, Eric Seidel, Tricia Ngoon, Adam Rule, AilieFraser, Benjamin Cosman, April Wang (masters)

Outreach@Scale My Python Tutor programming education website pythontutor.com has attractedover 3.5 million total users from over 180 countries. My personal website pgbovine.netcontains over 300 articles, videos, and podcast episodes on topics ranging from researchto education, and receives over 750,000 page views per year. I have also recorded over500 videos on research, education, and outreach topics for my YouTube channel, whichnow has 4,000+ subscribers and 500,000+ total video views.

Philip Guo – Page 11 of 11 – Curriculum Vitae

SANJOY DASGUPTA

Department of Computer Science and Engineering 9500 Gilman Drive #0404 La Jolla, CA 92093-0404

(858) 822-5270 dasgupta@cs.ucsd.edu

Research areas Machine learning, Algorithms Education Ph.D. in Computer Science, May 2000, University of California at Berkeley Thesis: Learning Probability Distributions Advisor: Umesh Vazirani B.A. in Computer Science, June 1993, Harvard University Thesis: Handwriting Recognition Using Hidden Markov Models Advisor: David Mumford Professional Experience Professor, Department of Computer Science and Engineering, University of California, San Diego July 2012 to Present Associate Professor, Department of Computer Science and Engineering, University of California, San Diego July 2008 to June 2012 Research Scientist, Machine Learning group, Yahoo! Labs January 2008 to March 2008 Assistant Professor, Department of Computer Science and Engineering, University of California, San Diego July 2002 to June 2008 Senior Technical Staff Member, Machine Learning group, AT&T Research Laboratory March 2000 to March 2002 Summer Researcher, Cryptography group, Bellcore Summer 1996 Researcher, Speech group, Bell Laboratories September 1993 to August 1994

Professional activities

Conference chairing Chair of organizing committee, Simons Institute program on Foundations of Machine Learning, Spring 2017 Program co-chair, International Conference on Machine Learning, 2013 Program co-chair, Conference on Learning Theory, 2009 Program co-chair, Workshop on Information Theory and Applications, 2009 Workshops co-chair, International Conference on Machine Learning (ICML), 2008 Local chair, Conference on Learning Theory (COLT), 2007 General co-chair, Workshop on Information Theory and Applications, 2007 Program co-chair, Workshop on Information Theory and Applications, 2006 Editorial boards and steering committees Action Editor, Journal of Machine Learning Research, 2006 to Present Associate Editor, Journal of Computer and System Sciences, 2007 to 2012 Board of directors, Association for Computational Learning (this is the steering committee for the COLT conference), 2004-2007 and 2018 to Present Member of the editorial board, Journal of Artificial Intelligence Research, 2003 to 2006 Member of the editorial board, Journal of Machine Learning Research, 2002 to 2005 Member of the editorial board, Machine Learning, 2000 to 2005 Conference organization and reviewing Area chair, Advances in Neural Information Processing Systems, 2007, 2008 Senior Program Committee, International Conference on Machine Learning, 2007, 2011 Senior Program Committee, Uncertainty in Artificial Intelligence, 2018 Program Committee, Conference on Artificial Intelligence (AAAI), 2007 Program Committee, Uncertainty in Artificial Intelligence, 2001, 2002, 2003, 2004, 2005, 2006 Program Committee, International Conference on Machine Learning, 2001, 2002, 2004, 2005, 2006, 2008 Program Committee, International Joint Conference on Artificial Intelligence, 2005 Program Committee, Advances in Neural Information Processing Systems, 2014, 2017 Program Committee, Conference on Learning Theory, 2004, 2008, 2014, 2015 Program Committee, Conference on Algorithmic Learning Theory, 2017 Senior Program Committee, Conference on Artificial Intelligence (AAAI), 2019 Program Committee, Conference on Innovations in Theoretical Computer Science, 2019

Funding history NSF grant CCF-1813160 “Algorithms for interactive learning”, dates: 6/2018-5/2021, amount: $500,000 (sole PI). NSF grant IIS-1237174 "Data e-platform leveraged for patient empowerment and population health improvement", dates: 10/2012-9/2016, amount: $2,000,000 (co-PI, with PI Kevin Patrick) NSF grant IIS-1162581 "Quantifying and utilizing confidence in machine learning", dates: 4/2012-3/2016, amount: $1,000,000 (co-PI, with PI Yoav Freund) NSF grant CNS-0932403 “CitiSense - Adaptive Services for Community-Driven Behavioral and Environmental Monitoring to Induce Change”, dates: 9/2010-8/2012, amount: $1,499,000 (co-PI, with PI Bill Griswold) NSF grant IIS-0812598 “Learning from data of low intrinsic dimension”, dates: 8/2008-7/2011, amount: $450,000 (co-PI, with PI Yoav Freund) NSF grant IIS-0713540 “Foundations of active learning”, dates: 8/2007 – 7/2010, amount: $428,000 (sole PI) NSF CAREER Award IIS-0347646 “Algorithms for unsupervised learning”, dates: 2/2004 – 1/2009, amount: $502,000 (sole PI) Awards 2013/14 Distinguished Teaching Award UC San Diego award for Senate members Teacher of the Year Award Computer Science Department, UC San Diego, 2011-12 Best Student Paper Award Experiments with Random Projection, Uncertainty in Artificial Intelligence, 2000 Outstanding Graduate Student Instructor Award University of California at Berkeley, 1999 and 2000

Invited Talks Annotation on the cheap Baidu Research, September 2018 Nearest neighbor methods Tutorial at International Conference on Machine Learning, July 2018 A neural algorithm for similarity search Invited talk, Symposium on Theory of Computing, June 2018 Interactive hierarchical clustering Baidu Research, September 2018 Summer School on Machine Learning, Los Alamos National Laboratory, June 2018 Using interaction for simpler and better learning ML seminar, University of Pennsylvania, April 2018 ML seminar, MIT, April 2018 Departmental colloquium, TTI-Chicago, April 2018 Interactive machine learning Osher Institute, May 2018 Learning from partial correction Colloquium, Department of Mathematics, University of Southern California, October 2017 Interactive structure learning Colloquium, Department of Computer Science, Duke University, November 2017 Colloquium, Electrical Engineering Department, University of California at Los Angeles, October 2017 Colloquium, Department of Operations Research and Financial Engineering, Princeton University, September 2017 Geometric algorithms for classification and retrieval in high dimension Symposium on Computational Geometry and Topology in the Sciences, College de France, June 2017 Cluster trees, near neighbor graphs, and continuum percolation Symposium on Geometry Understanding in Higher Dimensions, College de France, June 2017 Towards a theory of interactive learning Keynote talk, Conference on Artificial Intelligence and Statistics, April 2017 Embedding and projection approaches to fast nearest neighbor search Quarterly Theory Workshop, Northwestern, February 2017 Southern California Theory Day, Caltech, November 2016 Distribution-specific analysis of nearest neighbor search and classification Workshop on “Beyond worst-case complexity”, Simons Institute for Theoretical Computer Science, Berkeley, November 2016 A cost function for similarity-based hierarchical clustering

Workshop on Computational Challenges in Machine Learning, Simons Institute for Theoretical Computer Science, Berkeley, May 2017 Workshop on Foundations of Unsupervised Learning, Dagstuhl, Germany, September 2016 A protocol for interactive learning Southern California Symposium on Machine Learning, May 2016 ML distinguished lecture series, Carnegie Mellon University, May 2016 The geometry of interactive learning Workshop on Geometry and Machine Learning, Symposium on Theory of Computing, June 2016 Learning without supervision Talk at ViaSat, July 2016 Talk to UCSD alumni, May 2016 Lectures on clustering, active learning, and low intrinsic dimension Microsoft Research Summer School for Machine Learning, Bangalore, June 2015 Rates of convergence for nearest neighbor classification Symposium on Learning, Algorithms, and Complexity, Indian Institute of Science, January 2015 Fast algorithms for nearest neighbor search Information Sciences Institute, University of Southern California, July 2014 Computer Science Colloquium, UC Riverside, April 2014 Active learning and annotation Intelligent Systems Seminar, Los Alamos National Laboratory, July 2015 Duke Machine Learning Seminar, Duke University, April 2014 A short course in minimally supervised learning Series of three lectures at Conference on Structural Inference, Potsdam, September 2013 Parametrizing the easiness of machine learning problems Dagstuhl Workshop, “Analysis of Algorithms Beyond the Worst Case”, September 2014 Workshop on new theoretical directions in machine learning, Symposium on Theory of Computing, June 2013 Annotation on the cheap E-Harmony Labs, June 2012 ML distinguished lecture series, Carnegie-Mellon University, May 2012 Plenary talk, Snowbird workshop on machine learning, April 2012 Recent advances in active learning Workshop on machine learning in speech and language processing, International Conference on Machine Learning, 2011 A short course in unsupervised learning Series of three lectures at Institut Henri Poincare, Paris, May 2011

Cluster trees, near neighbor graphs, and continuum percolation ACM Colloquium, California Institute of Technology, February 2013 Computer Science Colloquium, University of Southern California, November 2012

Machine learning colloquium, Microsoft Research New England, February 2011 Plenary talk, Machine Learning Symposium, New York Academy of Sciences, October 2010 Nonparametric statistics in spaces of low intrinsic dimension Mathematics Colloquium, University of Southern California, September 2010 Linear-algebraic methods for learning latent variable models Tutorial at Conference on Uncertainty in Artificial Intelligence, Catalina Island, July 2010 The two faces of active learning Plenary talk, Summer Workshop on Speech Processing, Johns Hopkins University, July 2010 Keynote talk, Conference on Algorithmic Learning Theory, Porto, October 2009 Theory of active learning Tutorial at International Conference on Machine Learning, Montreal, June 2009 Analysis of clustering procedures Tutorial at Machine Learning Summer School, University of Chicago, June 2009 Two algorithms for hierarchical clustering Plenary talk, Conference of the International Federation of Classification Societies, March 2009 Learning from data of low intrinsic dimension Seminar, Center for Intelligent Systems, University of Alberta, November 2008 Computer Science Colloquium, University of California at Riverside, October 2008 Open problems in learning low dimensional representations Workshop on Geometry and Algorithms, Princeton, October 2008 Random projection trees and low dimensional manifolds Allerton Conference on Communication, Control, and Computing, Urbana-Champaign, September 2008 Workshop on Algorithms for Massive Data Sets, Stanford, June 2008 Workshop, Foundations of Computational Mathematics, City University of Hong Kong, June 2008 Computer Science and Statistics Joint Colloquium, Ohio State, May 2008 Seminar, Center for Information Systems, UC Berkeley, April 2008 Computer Science Colloquium, University of Southern California, April 2008 Machine Learning Seminar, New York University, March 2008 Seminar, AT&T Research Labs, March 2008 Algorithms Seminar, Duke University, February 2008 Workshop on Geometric and Topological Approaches to Data Analysis, University of Chicago, October 2007 Projection pursuit, Gaussian scale mixtures, and the EM algorithm Joint Statistical Meetings, Denver, August 2008 Seminar, IBM Research Labs, March 2008 Applied Math Colloquium, Columbia University, March 2008 Intelligence Seminar, Carnegie Mellon University, November 2007 Workshop on Geometry and Random Matrices, Statistical and Mathematical Sciences Institute, January 2007 Language/Learning/Vision/Graphics Seminar, MIT, November 2006 Active learning by reduction to supervised learning Seminar, Google Labs, January 2008

Seminar, Yahoo! Labs, January 2008 Workshop, International Symposium on Artificial Intelligence and Mathematics, January 2008 Active learning of linear separators Keynote talk, Annual meeting of Classification Society of North America, May 2006 Mathematics of Information Seminar, Caltech, February 2006 Biostatistics seminar, UC San Diego, October 2005 A central limit theorem for projections Seminar, Toyota Technological Institute in Chicago, May 2006 Models of active learning Workshop on Value of Information, NIPS, December 2005 Open problems in active learning Workshop on Foundations of Active Learning, NIPS, December 2005 Information geometry Machine Learning Summer School in Chicago, May 2005 Analysis of greedy active learning Seminar, Toyota Technological Institute, March 2005 Machine learning Tutorial, International Conference on Development and Learning, October 2004 Performance guarantees for hierarchical clustering UC Irvine AI and Statistics Seminar, October 2003 Seminar, Statistics department, University of Washington, December 2002 Seminar, Genome Institute of Singapore, June 2002 Learning mixtures of Gaussians Seminar, Information Sciences Institute, University of Southern California, December 2002 International Symposium on Mathematical Programming, August 2000 Algorithms for high-dimensional statistics Theory Seminar, IBM Almaden, April 2002 The complexity of learning Bayes nets Alpine Workshop on Machine Learning, April 2001 A probabilistic analysis of EM Alpine Workshop on Machine Learning, April 2001 KDD Seminar, IBM Research Labs, September 2000

Publications – Journals, Refereed Conferences, Books, and Book Chapters – reverse chronology

[C44] S. Dasgupta and C. Tosh Interactive structure learning via structural query-by-committee Advances in Neural Information Processing Systems, 2018 [C43] S. Dasgupta, A. Dey, N. Roberts and S. Sabato Learning from discriminative feature feedback Advances in Neural Information Processing Systems, 2018 [J16] C. Tosh and S. Dasgupta Maximum likelihood estimation for mixtures of spherical Gaussians is NP-hard Journal of Machine Learning Research, 18(175):1-11, 2018 [C42] E. Kazemi, L. Chen, S. Dasgupta and A. Karbasi Comparison based learning from weak oracles International Conference on Artificial Intelligence and Statistics, 2018 [J15] U. Mori, A. Mendiburu, S. Dasgupta and J.A. Lozano Early classification of time series by simultaneously optimizing the accuracy and earliness IEEE Transactions on Neural Networks and Learning Systems, 2017 [J14] S. Dasgupta, C.F. Stevens and S. Navlakha A neural algorithm for a fundamental computing problem Science, 358, 6364:793-796, 2017 [C41] C. Tosh and S. Dasgupta Diameter-based active learning International Conference on Machine Learning, 2017 [C40] S. Poulis and S. Dasgupta Learning with feature feedback International Conference on Artificial Intelligence and Statistics, 2017 [C39] X. Wang and S. Dasgupta An algorithm for L1 nearest neighbor search via monotonic embedding Advances in Neural Information Processing Systems, 2016 [C38] S. Vikram and S. Dasgupta Interactive Bayesian hierarchical clustering International Conference on Machine Learning, 2016 [C37] S. Dasgupta A cost function for similarity-based hierarchical clustering ACM Symposium on Theory of Computing, 2016 [J13] S. Dasgupta and K. Sinha Randomized partition trees for nearest neighbor search Algorithmica, 72(1):237-263, 2015

[C36] K. Chaudhuri and S. Dasgupta Rates of convergence for nearest neighbor classification Advances in Neural Information Processing Systems, 2014 [C35] S. Dasgupta and S. Kpotufe Optimal rates for k-NN density and mode estimation Advances in Neural Information Processing Systems, 2014 [C34] M. Ackerman and S. Dasgupta Incremental clustering: the case for extra clusters Advances in Neural Information Processing Systems, 2014 [J12] K. Chaudhuri, S. Dasgupta, S. Kpotufe, and U. von Luxburg Consistent procedures for cluster tree estimation and pruning IEEE Transactions on Information Theory, 60(12):7900-7912, 2014

[C33] C. Tosh and S. Dasgupta Lower bounds for the Gibbs sampler over Gaussian mixture models International Conference on Machine Learning, 2014 [C32] A. Balsubramani, S. Dasgupta, and Y. Freund The fast convergence of incremental PCA Advances in Neural Information Processing Systems, 2013 [C31] M. Telgarsky and S. Dasgupta Moment-based uniform deviation bounds for k-means and friends Advances in Neural Information Processing Systems, 2013 [C30] S. Dasgupta and K. Sinha Randomized partition trees for exact nearest neighbor search Conference on Learning Theory, 2013 [C29] S. Dasgupta Consistency of nearest neighbor classification under selective sampling Conference on Learning Theory, 2012 [C28] M. Telgarsky and S. Dasgupta Agglomerative Bregman clustering International Conference on Machine Learning, 2012 [J11] S. Kpotufe and S. Dasgupta A tree-based regressor that adapts to intrinsic dimension Journal of Computer and System Sciences, 78(5):1496-1515, 2012 [J10] S. Dasgupta

Two faces of active learning Theoretical Computer Science, 412(19):1767-1781, 2011 [BC1] S. Dasgupta

Theory of active learning Encyclopedia of Machine Learning, pp. 14-19, C. Sammut and G. Webb, eds, Springer, 2011

[C27] K. Chaudhuri and S. Dasgupta Rates of convergence for the cluster tree Advances in Neural Information Processing Systems, 2010 [J9] S. Dasgupta Strange effects in high dimension Communications of the ACM, 53(2):96, February 2010 [C26] N.A. Verma, S. Kpotufe, and S. Dasgupta Which spatial partition trees are adaptive to intrinsic dimension? 25th Conference on Uncertainty in Artificial Intelligence, 2009 [C25] A. Beygelzimer, S. Dasgupta, and J. Langford Importance weighted active learning International Conference on Machine Learning, 2009 [J8] S. Dasgupta and Y. Freund Random projection trees for vector quantization IEEE Transactions on Information Theory, 55(7), July 2009 [J7] S. Dasgupta, A. Kalai, and C. Monteleoni Analysis of perceptron-based active learning Journal of Machine Learning Research, 10:281-299, 2009 [C24] S. Dasgupta and D.J. Hsu Active learning by hierarchical sampling International Conference on Machine Learning, 2008 [C23] S. Dasgupta and Y. Freund

Random projection trees and low dimensional manifolds Fortieth ACM Symposium on Theory of Computing, 2008

[C22] S. Dasgupta, D.J. Hsu, and C. Monteleoni A general agnostic active learning algorithm Advances in Neural Information Processing Systems, 2007

[C21] L. Cayton and S. Dasgupta A learning framework for nearest-neighbor search Advances in Neural Information Processing Systems, 2007

[C20] Y. Freund, S. Dasgupta, M. Kabra, and N. Verma Learning the structure of manifolds using random projections Advances in Neural Information Processing Systems, 2007

[C19] S. Dasgupta and D.J. Hsu Online estimation with the multivariate Gaussian distribution Twentieth Annual Conference on Learning Theory, 2007

[J6] S. Dasgupta and L.J. Schulman A probabilistic analysis of EM for mixtures of separated, spherical Gaussians

Journal of Machine Learning Research, 8:203-226, 2007

[B1] S. Dasgupta, C.H. Papadimitriou, and U.V. Vazirani Algorithms McGraw-Hill, September 2006

[C18] S. Dasgupta, D.J. Hsu, and N.A. Verma A concentration theorem for projections Twenty-Second Conference on Uncertainty in Artificial Intelligence, 2006

[C17] L. Cayton and S. Dasgupta Robust Euclidean embedding International Conference on Machine Learning, 2006

[C16] S. Dasgupta Coarse sample complexity bounds for active learning Advances in Neural Information Processing Systems, 2005

[J5] T. Batu, S. Dasgupta, R. Kumar, and R. Rubinfeld The complexity of approximating the entropy SIAM Journal on Computing, 35(1), 2005.

[C15] A. Gupta and S. Dasgupta

Exploiting multiple views for hierarchical clustering Workshop on Multiple Views, International Conference on Machine Learning, 2005

[C14] S. Dasgupta, A. Kalai, and C. Monteleoni Analysis of perceptron-based active learning Eighteenth Annual Conference on Learning Theory, 2005

[J4] S. Dasgupta and P.M. Long Performance guarantees for hierarchical clustering Journal of Computer and System Sciences, 70(4): 555-569, 2005

[C13] S. Dasgupta Analysis of a greedy active learning strategy Advances in Neural Information Processing Systems 17, 2004

[C12] D. Kauchak and S. Dasgupta An incremental improvement procedure for hierarchical clustering Advances in Neural Information Processing Systems 16, 2003

[C11] S. Dasgupta and P.M. Long Boosting diverse weak learners Sixteenth Annual Conference on Computational Learning Theory, 2003

[J3] S. Dasgupta and A. Gupta An elementary proof of a theorem of Johnson and Lindenstrauss Random Structures and Algorithms, 22(1): 60-65, 2003

[J2] S. Dasgupta, W.S. Lee, and P.M. Long

A theoretical analysis of query selection for collaborative filtering Machine Learning, 51:283-298, 2003

[C10] T. Batu, S. Dasgupta, R. Kumar, and R. Rubinfeld The complexity of approximating the entropy Thirty-Fourth Annual ACM Symposium on Theory of Computing and Seventeenth IEEE Conference on Computational Complexity, 2002

[C9] S. Dasgupta, E. Pavlov, and Y. Singer An efficient PAC algorithm for reconstructing a mixture of lines Fourteenth Annual Conference on Algorithmic Learning Theory, 2002

[C8] S. Dasgupta Performance guarantees for hierarchical clustering Fifteenth Annual Conference on Computational Learning Theory, 2002

[C7] S. Dasgupta, M.L. Littman, and D. McAllester PAC generalization bounds for co-training Advances in Neural Information Processing Systems 14, 2001

[C6] M. Collins, S. Dasgupta, and R.E. Schapire A generalization of principal component analysis to the exponential family Advances in Neural Information Processing Systems 14, 2001

[C5] D. Precup, R. Sutton, and S. Dasgupta Off-policy temporal difference learning with function approximation Eighteenth International Conference on Machine Learning, 2001

[C4] S. Dasgupta and L.J. Schulman A two-round variant of EM for Gaussian mixtures Sixteenth Conference on Uncertainty in Artificial Intelligence, 2000

[C3] S. Dasgupta Experiments with random projection Sixteenth Conference on Uncertainty in Artificial Intelligence, 2000

[C2] S. Dasgupta Learning polytrees Fifteenth Conference on Uncertainty in Artificial Intelligence, 1999

[C1] S. Dasgupta Learning mixtures of Gaussians Fortieth Annual IEEE Symposium on Foundations of Computer Science, 1999

[J1] S. Dasgupta

The sample complexity of learning fixed-structure Bayesian nets Machine Learning, 1997

Advisees Alumni Dustin Boswell, M.Sc., 2004 Techniques for spelling correction

Software engineer at Google

Anjum Gupta, M.Sc., 2005 Data mining under abstract distance and projection operators

Small business owner

Konstantinos Bimpikis, M.Sc., 2005 Associate Professor, Stanford Huayong Hu, M.Sc., 2005 Now a physician

Claire Monteleoni, postdoc, 2006-8 Associate Professor, George Washington University Lawrence Cayton, Ph.D., 2009 Algorithms for proximity search and low-dimensional embedding Software engineer Kamalika Chaudhuri, postdoc, 2009-10 Associate Professor, CSE department, UC San Diego Daniel Hsu, Ph.D., 2010 Algorithms for active learning Assistant Professor, Columbia University Samory Kpotufe, Ph.D., 2010 Learning from data of low intrinsic dimension Assistant Professor, Princeton University Kaushik Sinha, postdoc, 2010-12 Assistant Professor, Wichita State University Nakul Verma, Ph.D., 2012 Learning from data with low intrinsic dimension

Lecturer, Columbia University Matus Telgarsky, Ph.D., 2013 Duality and data dependence in boosting Assistant Professor, University of Illinois at Urbana-Champaign Margareta Ackerman, postdoc, 2013-14 Assistant Professor, Santa Clara University Christopher Tosh, Ph.D., 2018

Algorithms for sampling and interactive learning Now postdoc, Columbia University

Current students Stefanos Poulis Interactive learning

Sharad Vikram Interactive learning Andrew Leverentz Interactive scientific discovery

UNIVERSITY OF CALIFORNIA, SAN DIEGO

BERKELEY • DAVIS • IRVINE • LOS ANGELES • RIVERSIDE • SAN DIEGO • SAN FRANCISCO SANTA BARBARA • SANTA CRUZ

PROFESSOR SHLOMO DUBNOV (858) 534-5941 DEPARTMENT OF MUSIC (0326) sdubnov@ucsd.edu 9500 GILMAN DRIVE FAX: (858) 534-8502 LA JOLLA, CALIFORNIA 92093-0326

DUBNOV, Shlomo short CV

Previous Applicable Employment 2016: Affiliate Professor, University of Haifa, Israel

2013: Affiliate Professor, Computer Science and Engineering, UCSD

2011 - today: Full Professor, Music Department, UCSD

2005 – 2011: Associate Professor, Music Department, UCSD

2003 – 2005: Assistant Professor, Music Department, UCSD

1998 – 2003: Senior Lecturer, Department of Communication Systems Engineering, Ben-Gurion University,

Israel,

1998 - 1999: Lecturer, Music Department, Bar-Ilan University.

1996 - 1998: Invited Researcher, IRCAM, Paris.

1994 - 1996: Lecturer, the Jerusalem Rubin Academy of Music and Dance

1989 - 1996: Consultant, Department of Operations Research, Israel Ministry of Defense.

1983 - 1988: Project Officer, ``RAFAEL'', Research Institute of the Israel Ministry of Defense.

Education

B.Sc: Physics/Mathematics, Hebrew University, 1980-1983

M.Sc: Electrical Engineering, Technion, Israel Institute of Technology, Haifa., 1985-1988,

Private composition studies with Zvi (Harry) Nadel, 1985-1987

B.Mus: Theory and Composition, The Jerusalem Rubin Academy of Music and Dance, 1988-1991,

Ph.D: Computer Science in cooperation with Musicology, The Hebrew University of Jerusalem, 1991-1996,

Post Doctorate: Analysis/Synthesis Team, IRCAM, Centre Pompidou, Paris, 1996-1997

Research Administration:

Director, Center for Research in Entertainment and Learning, Qualcomm Institute, CALIT2, UCSD

Founding faculty, Computer Audition Lab, CALIT2, UCSD

Research Partner, DYCI2 : Dynamiques créatives de l'interaction improvisée, ANR, France

Selected Recent Professional functions outside universities/institutions 2008 - Senior Member IEEE

2009 - Secretary of IEEE Technical Committee on Computer Generated Music

2012 - Associate Editor, Computers in Entertainment, ACM (http://cie.acm.org/about-us/)

2013 - Program co-chair, IEEE Conference on Semantic Computing, Irvine, CA, Sep 2013

2014 - Visiting Professor, MDBL Lab, KEIO University, Japan

2016 - Visiting Professor, LaBRI, University of Bordeaux, IdEX Excellence in Research Fellowship, France

Recent Invited Talks and Performances

- “Discrete Musical Systems”, series of 4 talks in El Centro Latinoamericano de Formación Interdisciplinaria

(CELFI), Buenos Aires, 2016

- “Composer Duets” algorithmic composition, Carlos Britez and Maria Amarilla (violins), Buenos Aires, 2016

- “Information Theoretic Creativity”, Keynote talk, Journées d’Informatique Musicale, 2016, Albi, France

- “Algorithmic Music: Creativity, Genius, and AI”, talk and concert, with Geoffrey Keezer (piano), Mozart and the

Mind Symposium 2015, San Diego

- “Machinations2”, machine improvisation with Finn Peters (flute), MuMe 2014, Café Oto, London

- “The Science of Story”, talk at StoryExpo, 2013, Hollywood,

- “Tutorial on Semantic Computing for Edutainment”, IEEE ICSC, 2013 Irvine, USA,

- “Live Data Visualization: You've Got Data -- Now What?”, Panel at Digital Hollywood 2013

- “Nomos Machina” performance at ISEA, Mume 2013, Sydney, talk and performance at IRCAM, Conference on

Geometric Sciences of Information 2013, Paris

Published/Creative Work

Four main thrusts can be identified in my work: (1) machine improvisation and statistical modeling of musical

style; (2) analysis of musical timbre; (3) cognitive modeling of music and musical anticipation; and (4) music

composition and creative production. My research appears in technical and scientific journals (JASA; IEEE

Transactions; Signal Processing) as well as more specialized journals considered the flagships of the computer

music field (Computer Music Journal; Journal of New Music Research; Organised Sound). I am one of the

founders of the field of machine improvisation where I contributed novel machine-learning algorithms for capturing

musical style. I’m also an initiator of a new research area of Musical Information Dynamics that uses information

theoretical tools for modeling cognitive and anticipatory aspects of musical structure. I’ve been working on

number of topics related to Computing in Entertainment and Computational Creativity. Recordings of my creative

works were published by major music publishers and aired nationwide on UCSD TV and JLTV. I served as a

secretary of IEEE Technical Committee on Computer Generated Music, and currently act as a lead editor in ACM

Computers in Entertainment. In terms of research management, I’ve been active in directing broad range of

interdisciplinary research at the Center for Research in Entertainment and Learning (CREL) in the Qualcomm

Institute in UCSD (Calit2/QI). CREL brings together faculty from Engineering, Social Sciences and Humanities to

create a unique environment for nurturing novel research ideas and developing practical entertainment and

learning applications. My publications include over 100 peer-reviewed articles, co-authoring of two books and

several dozens of artistic works. My Google Scholar h-index is 25, and Research Gate Score of 24.96, which is

among the very top computer music researchers worldwide.

OMB No. 0925-0001 and 0925-0002 (Rev. 10/15 Approved Through 10/31/2018)

BIOGRAPHICAL SKETCH Provide the following information for the Senior/key personnel and other significant contributors.

Follow this format for each person. DO NOT EXCEED FIVE PAGES.

NAME: Bradley Voytek eRA COMMONS USER NAME (credential, e.g., agency login): bvoytek POSITION TITLE: Associate Professor EDUCATION/TRAINING (Begin with baccalaureate or other initial professional education, such as nursing, include postdoctoral training and residency training if applicable. Add/delete rows as necessary.)

INSTITUTION AND LOCATION

DEGREE (if

applicable)

Completion Date

MM/YYYY

FIELD OF STUDY

University of Southern California (Los Angeles, CA, USA) University of California, Berkeley (CA, USA) University of California, Berkeley (CA, USA) University of California, San Francisco (CA, USA)

B.A. Ph.D. Post-doc Post-doc

05/2002 05/2010 05/2011 02/2014

Psychology Neuroscience Neuroscience Neuroscience

A. Personal Statement During the previous 8 years, my NIH-funded research—beginning with a PhD Diversity Supplement to parent R01 followed by a post-doctoral Institutional Research�and Career Development (diversity) Award—has focused on the roles that neural oscillations play in human cognition. While those projects were performed using only human electrophysiological recordings, my current research has expanded to involve detailed computational modeling, neural simulation, and animal models. These new research paths are focused on understanding the physiological origins of neural oscillations and spectral features, the role oscillations play in neural computation and/or coordinating information transfer between brain regions, and how they influence human cognition and cognitive dysfunction. To do this, my research program combines hypothesis-driven experimental cognitive neuroscience, computational neuronal modeling, large-scale data mining, and machine learning. Our goal is to identify the mechanisms for neural communication, with a focus on the role that neuronal excitation/inhibition plays in neural coding and the formation of oscillations. Although aimed at normal human cognition—including working memory and attention—my program also examines how disruptions to these processes underlie cognitive problems associated with healthy aging as well as neurological and psychiatric diseases. I have broad experience in computational, systems, and cognitive neuroscience, making use of invasive electrocorticography in patients undergoing brain surgery for intractable epilepsy, non-invasive scalp M/EEG, as well as invasive and non-invasive brain stimulation. I have the experience and training necessary for successful completion of the proposed research project. Although I am a new investigator, I have quickly established my lab and continue a number of fruitful collaborations with senior, tenured, NIH-funded scientists. During 2011-2012 my academic career was on hold as I worked full-time at a technology startup before returning to academia. During that “startup sabbatical” year, I gained significant training and education in software development and programming, machine learning, and data science. I also took time off (“active service, modified duty” at UCSD) from 2013 to early 2014 to care for my baby daughter. I fully resumed my research by mid-2014. In addition to my research, I am actively involved in university service and outreach/mentoring. I am on the UCSD Neurosciences Graduate Program Executive Committee, where I also serve as Chair of the Diversity Committee. I am also a founding member and Executive Board Member of the UCSD Data Science major and Halıcıoğlu Data Science Institute, and the UCSD Kavli Institute for Brain and Mind, the latter of which I also serve as an Advisory Board Member.

B. Positions and Honors Positions and Employment 2011 - 2014 Data Scientist, Uber, Inc. 2014 - 2018 Assistant Professor, University of California, San Diego 2017 - Executive Committee, Halıcıoğlu Data Science Institute, University of California, San Diego 2017 - Executive Committee, Neurosciences Graduate Program, University of California, San Diego 2018 - Associate Professor, University of California, San Diego Other Experience and Professional Memberships 2004 - Society for Neuroscience 2012 - American Cognitive Neuroscience Society Academic and Professional Honors 2005: University of California, Berkeley Outstanding Graduate Instructor Award 2007: University of California Berkeley, Course Improvement Grant, Functional Neuroanatomy 2009: American Association for the Advancement of Science, Program for Excellence in Science award 2009: Society for Neuroscience, Neuroscience Scholars Program fellow 2011: Consultant, The National Academy of Sciences - Science & Entertainment Exchange 2012: University of California, San Francisco. Information Technology Innovation�Contest winner 2012: AAAS: Finalist - Early Career Award for Public Engagement with Science 2013: University of California: President's Postdoctoral Fellowship Program 2015: Alfred P. Sloan Foundation Neuroscience Research Fellowship 2015: National Academy of Sciences Kavli Fellow 2016: Computational and Systems Neuroscience New Attendee Award 2018: Fellow: Halıcıoğlu Data Science Institute (HDSI) of the University of California, San Diego Service 2014 - 2016 UC San Diego, Neurosciences Graduate Program: Diversity Committee 2014 - 2017 UC San Diego, Department of Cognitive Science: Faculty Diversity Representative 2014 - 2017 UC San Diego, Cognitive Science Student Association: Faculty Representative 2015 NIH/CSHL Banbury: “Brain Rhythms as Potential Targets for Intervention in Cognitive

Dysfunctions” 2015 - 2017 UC San Diego, John Muir College: Academic Senate Representative (elected) 2015 - 2017 National Academy of Sciences Kavli Frontiers of Science symposium, Organizing

committee 2015 - current UC San Diego, Data Science Student Society (DS3): Faculty Representative 2016 - current UC San Diego, Data Science undergraduate major: Founding Faculty 2016 - current Kavli Institute for Brain and Mind: Advisory Board 2016 - current UC San Diego, Neurosciences Graduate Program: Diversity Chair 2017 - Diversity Outreach and Training Committee founder and member: Cognitive Neuroscience

Society 2017 - 2018 National Academy of Sciences Kavli Frontiers of Science symposium: Chair 2017 - current UC San Diego, Halıcıoğlu Institute for Data Science: Founding faculty and Executive Board 2017 - current Executive Committee: Neurosciences Graduate Program, UC San Diego 2017 - current Executive Committee: Halıcıoğlu Data Science Institute, UC San Diego 2018 - Director: Halıcıoğlu Data Science Institute Undergraduate Research Fellowship Program,

UC San Diego 2018 - Director: Halıcıoğlu Data Science Institute Distinguished Lecture Series, UC San Diego C. Contributions to Science 1. Focal brain lesions impact neural information flow: Working with patients with focal brain lesions, in my

PhD research I used scalp EEG to show how prefrontal cortex (PFC) and basal ganglia (BG) affect top-down visual attention and working memory (a). Furthermore, I demonstrated that, although significantly

detrimental, PFC and BG lesions do not abolish visual working memory or attention. Rather, these cognitive deficits are incomplete. We showed that successful cognitive performance can be partially explained by a dynamic process whereby the intact, un-lesioned PFC compensates as the lesioned PFC is challenged (b). We showed that we can “block” this compensation by adding visual noise during the precise times when visual information should cross interhemispherically from the damaged to non-damaged hemisphere (c).

a. Voytek B & Knight RT. Prefrontal cortex and basal ganglia contributions to visual working memory. Proc Natl Acad Sci USA 107(42), 18167-18172, 2010.

b. Voytek B, Davis M, Yago E, Barceló F, Vogel EK, Knight RT. Dynamic neuroplasticity after human prefrontal cortex damage. Neuron 68(3), 401-408, 2010.

c. Voytek B*, Soltani M*, Pickard N, Kishiyama MM, Knight RT. Prefrontal cortex lesions impair object-spatial integration. PLoS ONE 7(4), e34937, 2012.

d. Løvstad M, Funderud I, Lindgren M, Endestad T, Due-Tønnessen P, Meling T, Voytek B, Knight RT, Solbakk AK. Contribution of subregions of human frontal cortex to novelty processing. J Cogn Neurosci 24(2), 378-395, 2012.

2. Phase-amplitude coupling as cortical communication mechanism: Building off my lesion EEG

research, I began to explore the role that neural oscillations play in cognition and in coordinating information transfer between brain regions. To do this, I began to use invasive electrocorticography in patients undergoing surgery for intractable epilepsy. To address my cognitive questions of interest, I had to develop and validate several novel methods and procedures: one outlines the rationale for using electrocorticography and quantifies the signal-to-noise improvement over scalp EEG (a); another introduces a novel metric for statistically assessing the relationship between low-frequency oscillations and local population spiking activity, known as phase-amplitude coupling (PAC) (b). Taking advantage of the improved signal-to-noise of invasive human electrocorticography (ECoG), I first showed that activity in different cortical regions preferentially couples to different oscillatory frequencies (c). I then incorporated all of these methods and ideas, and combined them with a sophisticated behavioral paradigm to show how oscillatory dynamics play a critical role in coordinating cognitive networks during human goal maintenance (d). We found that oscillatory networks for quickly, with different task goals represented at different coupling phases to permit directional information transfer on a trial-by-trial basis.

a. Voytek B, Secundo L, Bidet-Caulet A, Scabini D, Stiver S, Gean AD, Manley G, Knight RT. Hemicraniectomy: A new model for human electrophysiology with high spatio-temporal resolution. J Cogn Neurosci 22(11), 2491-2502, 2010.

b. Voytek B, D’Esposito M, Crone NE, Knight RT. A method for event-related phase/amplitude coupling. NeuroImage 64, 416-424, 2013.

c. Voytek B, Canolty RT, Shestyuk A, Crone NE, Parvizi J, Knight RT. Shifts in gamma phase-amplitude coupling frequency from theta to alpha over posterior cortex during visual tasks. Front Hum Neurosci 4(191), 1-9, 2010.

d. Voytek B, Kayser A, Badre D, Fegen D, Chang EF, Crone NE, Parvizi J, Knight RT, D’Esposito M. Oscillatory dynamics coordinating human frontal networks in support of goal maintenance. Nature Neurosci, 2015.

3. Periodic and aperiodic origins of cognition and cognitive decline in health and disease: During my

post-doctoral fellowship I sought to extend my work on focal brain lesions and oscillation-mediated cognitive functioning to healthy, non-pathological domains. Here I focused on the “neural noise” theory of age-related cognitive decline to ground this theory in physiology. We showed that older adults have both decreased PAC and increased 1/f “neural noise,” and that this 1/f noise can be observed in scalp EEG, and mediates age-related visual working memory decline (a). In my lab, we followed up on this by looking at scalp EEG measures of visual cortical alpha phase dynamics, and found that failures in alpha phase alignment to an alerting cue propagate out a second later resulting in working memory impairments (b). Interestingly, the multivariate pattern of visual cortical alpha activity appears to encode spatial focus of attention (c). We have recently showed that instead of irregularities in many frequency bands, 1/f neural noise may be a better biomarker for schizophrenia (d). Finally, my lab has introduced new tools for parameterizing neural power spectra into their component periodic and aperiodic (1/f) components, including online tutorials (e).

a. Voytek B, Kramer MA, Case J, Lepage KQ, Tempesta ZR, Knight RT, Gazzaley A. Age-related changes in 1/f neural electrophysiological noise. J Neurosci, 2015.

b. Tran TT, Hoffner NC, LaHue SC, Tseng L, Voytek B. Alpha phase dynamics predict age-related visual working memory decline. NeuroImage 143, 196-203. 2016.

c. Voytek B, Samaha J, Rolle CE, Greenberg Z, Gill N, Porat S, Kader T, Rahman S, Malzyner R, Gazzaley A. Preparatory encoding of the fine scale of human spatial attention. J Cogn Neurosci. 2017.

d. Peterson EJ, Rosen BQ, Campbell AM, Belger A, Voytek B. 1/f neural noise is a better predictor of schizophrenia than neural oscillations. (Pre-print on bioRxiv; manuscript under review).

e. Haller M*, Donoghue T*, Peterson E*, Varma P, Sebastian P, Gao R, Noto T, Knight RT, Shestyuk A#, Voytek B#. Parameterizing neural power spectra. (Pre-print on bioRxiv; manuscript in revision).

4. Excitation/inhibition (EI) balance and cognition: Drawing on the results that 1/f neural noise is associated with cognitive impairments, my lab has begun work exploring the more precise neural origins of changes in the 1/f power spectrum. Specifically, we have been exploring the role that excitation/inhibition (EI) balance plays in oscillatory formation, information transfer, and neural coding. This culminated in my development of a novel theoretical framework whereby EI imbalances impact neural spiking and information transmission (a). Although this paper is only been out for about 28 months, it has already been influential (70 citations). After presenting some of our recent work in this domain, I was invited to write a Full Review regarding the physiological information that can be extracted from EEG (b). Next, using a mix of computational modeling as well as data from rat hippocampal recordings and macaque cortex, we demonstrate that this 1/f “neural noise” may index E:I balance (c). Finally, we implemented a novel model of alpha oscillations as a rhythmic modulation of background EI noise, which predicted that alpha exists in two modes: bursting, which serves to enhance gain of the neuronal population and sustained, which suppresses gain; these predictions were then borne out in human physiology (d).

a. Voytek B, Knight RT. Dynamic network communication as a unifying neural basis for cognition, development, aging, and disease. Biol Psychiatry 77(12), 1089-1097, 2015.

b. Voytek B, Postle B, Watrous A, van der Meij R, Gao RD, Peterson EJ. Inferring neurophysiology and network-level dynamics from the human EEG. (Invited Perspective in revision at Nature Neurosci)

c. Gao RD, Peterson EJ, Voytek B. Inferring synaptic excitation/inhibition balance from field potentials. NeuroImage 158, 70-78, 2017.

d. Peterson EJ & Voytek B. Alpha oscillations control cortical gain by modulating excitatory-inhibitory background activity. (Pre-print on bioRxiv; manuscript in revision).

5. How neural oscillation waveform shape informs physiology: My lab has recently pioneered an entirely

new physiological framework for analyzing the shape of neural oscillations, showing that non-sinusoidal features may reflect alterations in synaptic input in e.g., Parkinson’s disease (a). Additionally we have introduced a novel suite of tools for analyzing the shape of local field potential and EEG waveforms, which we have argued allow for better inference of the physiology of the field potential generators and how they are influenced by cognitive, behavioral, and disease states (b). Both papers have already been quite influential, receiving 19 and 43 citations, respectively, despite having only been out for just over a year. Finally, we have introduced a suite of Python tools—tested and validated against simulated ground-truth and animal data—for performing these analyses, including full tutorials(c).

a. Cole SR, van der Meij R, Peterson EJ, de Hemptinne C, Starr PA, Voytek B. Nonsinusoidal oscillations underlie pathological phase-amplitude coupling in the motor cortex in Parkinson's disease. J Neurosci 37(18), 4830-4840. 2017.

b. Cole SR & Voytek B. Brain oscillations and the importance of waveform shape. Trends Cogn Sci 21, 140-152. 2017.

c. Cole SR & Voytek B. Cycle-by-cycle analysis of neural oscillations. (Pre-print on bioRxiv; manuscript in revision).

6. Data analytics: Finally, my research has begun to make extensive use of large-scale data analysis and I’m moving my research lab to an “open science” format, including the publishing of available data (when appropriate and authorized) as well as publishing all analysis code using free and open software (e.g., Python). I’ve demonstrated how data mining of the peer-reviewed neuroscientific literature can aid research and scientific discovery (a). In an invited commentary, I argue the advantages of creating an open data

ecosystem for facilitating these kinds of data-mining projects (b) and (c). All of my methods and data (where possible) are made public, and my lab has been developing interactive tutorials and publishing software packages made freely available to everyone on GitHub (d).

a. Voytek JB, Voytek B. Automated cognome construction and semi-automated hypothesis generation. J Neurosci Methods 208(1), 92-100, 2012.

b. Voytek B. The virtuous cycle of a data ecosystem. PLoS Comput Biol 12(8), 1-6, 2016. c. Voytek B. The Inextricable Links between Open Science, Data Science, and Social Media. Neuron

96, 1219-1222, 2017. d. https://github.com/voytekresearch

Complete List of Published Work in MyBibliography: http://www.ncbi.nlm.nih.gov/sites/myncbi/bradley.voytek.1/bibliography/40243234/public/?sort=date&direction=descending D. Research Support Active Research Support 2017 – 2020 NSF BCS COGNEURO 1736028: Attention, Oscillatory phase dynamics coordinate cognitive

neural networks, $471,775 (total). (Sole-PI: B. Voytek) 2017 – 2020 NSF DGE NRT 1735234 NRT-IGE: Augmenting, Piloting, and Scaling Computational

Notebooks to Train New Graduate Researchers in Data-Centric Programming, $500,000 (total). (Co-PI; PI: James Hollan; Co-PIs: Philip Guo, Scott Klemmer)

2018 – 2020 Whitehall Foundation, $225,000 (total). (Sole-PI: B. Voytek) Completed Research Support 2008 – 2010 National Institute of Neurological Disorders and Stroke, NIH. Diversity Supplement to

NS021135: Attention, Orientation and Human Prefrontal Cortex, $122,988. (Graduate student researcher; PI: RT. Knight)

2011 – 2014 National Institute of General Medical Sciences, NIH. Institutional Research�and Career Development Award (IRACDA) Scholars in Science (ISIS), ~$300,000 (four years salary and benefits, plus research and travel costs). (Mentor: A. Gazzaley)

2014 – 2015 UCSD Qualcomm Institute (QI), California Institute for Telecommunications and Information Technology (Calit2) Strategic Research Opportunities (CSRO) program, $50,000

2015 Kavli Institute for Brain and Mind Innovative Research Grant: Finding parallels: the role of the subiculum in neural encoding of object-environment alignment, $50,000 (Co-PI; PIs: J. Olson)

2015 – 2016 Alfred P. Sloan Research Fellow in Neuroscience, $50,000 2015 – 2016 R01 MH095984-03S1, Revision Application (formerly "Supplement") to parent R01 "Oscillatory

Contributions to Working Memory and Attention" (Co-Investigator; PI: B.R. Postle), Total direct costs $133,581

2016 Kavli Institute for Brain and Mind Innovative Research Grant: Systems- and Synaptic-Level Overcoupling in Major Depressive Disorder, $50,000 (Co-PI; PI: T. Tran)

2017 Kavli Institute for Brain and Mind Innovative Research Grant: Bridging Structure and Function with Neural Oscillations in iPSC-derived Cortical Organoids, $50,000 (Co-PI; PI: T. Olayinka)

Yannis Papakonstantinou is a Professor of Computer Science and Engineering at the

University of California, San Diego. He has published over 110 research articles that have

received more than 15,000 citations, according to Google Scholar. A common theme of his

research is the extension of database platforms and query processors beyond centralized

relational databases and into semistructured databases, integrated views of distributed

databases and web services, textual data and queries involving keyword search, and most

recently spatiotemporal sensor data. He has given multiple tutorials and invited talks, has

served on journal editorial boards and has chaired and participated in program committees

for many international conferences and workshops. He was a co-director of UCSD's Master

of Advanced Studies in Data Science from its founding until 2018.

Yannis enjoys to commercialize his research and to inform his research accordingly. He was

the CEO and Chief Scientist of Enosys Software, which built and commercialized an early

Enterprise Information Integration platform for structured and semistructured data. The

Enosys Software was OEM'd and sold under the BEA Liquid Data and BEA Aqualogic brand

names, eventually acquired in 2003 by BEA Systems. His lab's FORWARD platform was used

by many UCSD and commercial applications. He has been involved in data analytics in the

pharmaceutical industry, was in the technical advisory board of Brightscope Inc and

TigerGraph Inc, and he has been working with Amazon Web Services. He is the inventor of

ten patents.

Yannis holds a Diploma of Electrical Engineering from the National Technical University of

Athens, MS and Ph.D. in Computer Science from Stanford University (1997) and an NSF

CAREER award for his work on data integration.

Yoav S. Freundyfreund@ucsd.edu

www.cse.ucsd.edu/∼yfreund

University of California, San DiegoComputer Science and Engineering9500 Gilman Drive #0404La Jolla, California 92093-0404(858) 534-1668

DegreesB.Sc. (Physics and Mathematics) 1982, Hebrew University in Jerusalem.M.Sc. (Computer Science) 1989, Hebrew University in Jerusalem.

Thesis: The Role of Quantization in Learning Algorithms, supervised by Eli ShamirPh.D. (Computer Science) 1993, University of California at Santa Cruz.

Thesis: Data Filtering and Distribution Modeling Algorithms for Machine Learning, supervisedby Manfred Warmuth and David Haussler.

HonorsThe 2003 Godel prize for the paper “A decision-theoretic generalization of on-line learning and an

application to boosting” by Freund and Schapire.

The 2004 Paris Kanellakis Theory and Practice Award (ACM).

Fellow of the American Association for Artificial Intelligence (AAAI) since 2008.

Listed in isihighlycited.com as one of the 339 most cited researchers in computer science.

Professional Experience

1982 to 1985 Israel Armament Development Authority.Image processing laboratory.Developed image processing and pattern recognition algorithms.

1985 to 1989 Israel Military Industries.Advanced systems division.Headed a five person software team developing a real-time control system.

1993 to 2001 AT&T Bell Laboratories, later AT&T LabsMachine learning group.Member of technical staff.

2001 to 2003 Banter Inc.Machine learning expert and algorithm developer.

2003 to 2005 Columbia UniversitySenior Research Scientist.The Center for Computational Learning Systems.

Sept. 2005 to now University of California, San DiegoProfessor of Computer Science.

BookRobert E. Schapire and Yoav Freund. Boosting, Foundations and Algorithms. MIT Press. 2012.

Online CourseUCSD/edX micro-masters in Data Science. Leading the four-course series and creating the online

course “Big Data Analytics using Spark”.https://www.edx.org/micromasters/data-science

PublicationsChen,Y. McElvain, L., Tolpygo, A. Ferrante, D. Karten, H. Mitra, P. Kleinfeld, D., Freund, Y.. The

active atlas: Combining 3D anatomical models with texture detectors. Medical Image Computingand Computer Assisted Intervention MICCAI 2017. Lecture Notes in Computer Science, vol10433 (2017)

Katsis, Yannis, Yoav Freund, Yannis Papakonstantinou. Combining Databases and Signal Processingin Plato. Conference on Conference on Innovative Data Systems Research CIDR. 2015.

Akshay Balsubramani, Yoav Freund. Scalable Semi-Supervised Aggregation of Classifiers. NeuralInformation Processing Systems (NIPS), 2015.

Akshay Balsubramani, Yoav Freund, Optimally Combining Classifiers Using Unlabeled Data. Con-ference on Learning Theory (COLT) 2015.

Matthew Elkherj, Yoav Freund. A System for Sending the Right Hint at the Right Time. ACM:Learning at Scale 2014 March 2014.

Akshay Balsubramani, Sanjoy Dasgupta, Yoav Freund, The Fast Convergence of Incremental PCA.Neural Information Processing Systems 2013.

Valmianski I, Shih AY, Driscoll JD, Matthews DW, Freund Y, Kleinfeld D. Automatic identificationof fluorescently labeled brain cells for rapid functional imaging. Journal of NeurophysiologySeptember 2012

Nathan Jacobson, Yoav Freund, Truong Nguyen, An Online Learning Approach to Occlusion Bound-ary Detection, IEEE Transaction on Image Processing January 2012

Matthew Jacobsen, Yoav Freund, Ryan Kastner, RIFFA: A Reusable Integration Framework for FPGAAccelerators. 20th Annual International Symposium on Field-Programmable Custom ComputingMachines (FCCM), May 2012.

Mayank Kabra, (Advisor Yoav Freund) Automated cancer detection and drug discovery: two biomed-ical vision systems. PhD. Thesis 2011.

Evan Ettinger and Yoav Freund, Camera Pointing with Coordinate-Free Localization and Tracking,Advances in Sound Localization Pawel Strumillo (Ed.) 2011.

Kamalika Chaudhuri, Yoav Freund, Daniel Hsu. Tracking using explanation-based modeling. Uncer-tainty in AI 2010.

Kamalika Chaudhuri, Yoav Freund, Daniel Hsu. A parameter-free hedging algorithm. Neural Infor-mation Processing Systems 2009.

Sanjoy Dasgupta and Yoav Freund, Random projection trees for vector quantization. IEEE Transac-tions on Information Theory July 2009, Vol 55, Issue 7.

Yoav Freund. A method for Hedging in continuous time. May 2009. Arxiv/0904.3356

2

Valmianski I, Shih AY, Driscoll JD, Matthews DW, Freund Y and Kleinfeld D. Automatic identifi-cation of fluorescently labeled brain cells for rapid functional imaging. J Neurophysiol. 2010Sep;104(3):1803-11. 2.

Arvey A, Hermann A, Hsia CC, Ie E, Freund Y and McGinnis W. Minimizing off-target signals inRNA fluorescent in situ hybridization. Nucleic Acids Res. 2010 Jun;38(10)

Par A, Lemons D, Kosman D, Beaver W, Freund Y, McGinnis W. Visualization of individual ScrmRNAs during Drosophila embryogenesis yields evidence for transcriptional bursting. CurrentBiology 2009 Dec 15;19(23):2037-42.

Alterovitz R, Arvey A, Sankararaman S, Dallett C, Freund Y, Sjolander K. ResBoost: characterizingand predicting catalytic residues in enzymes. BMC Bioinformatics 2009 Jun 27;10(1):197.

Evan Ettinger and Yoav Freund. Coordinate-Free Calibration of an Acoustically Driven Camera Point-ing System. International Conference on Distributed Smart Cameras (ICDSC), September 2008.

Roy Liu, Yoav Freund and Glen Spraggon, Image-based crystal detection: a machine-learning ap-proach, ActaCryst D. Biological Crystalography, Vol 64, No 12, 2008.

mayank Kabra, Yoav Freund and Stephen Baird, Selective Scanning for faster Prostate Pathology.Workshop on Bio-Image Informatics 2008

William Beaver, Adam Pare, David Kosman, Ethan Bier, William McGinnis and Yoav Freund Collec-tions of classifiers tuned for cell finding within and application to building digital cell atlases ofDrosophila embryos. Workshop on Bio-Image Informatics 2008

Yoav Freund, Sanjoy Dasgupta, Mayank Kabra and Nakul Verma Learning the structure of manifoldsusing random projections Neural Information Processing Sysems 2007

William Beaver, David Kosman, Gary Tedeschi, Ethan Bier, William McGinnis, Yoav Freund. Seg-mentation of Nuclei in Confocal Image Stacks Using Performance-based Thresholding. Proceed-ings of the IEEE International Symposium on Biomedical Imaging (ISBI), 2007.

Gregory Giannone, Benjamin J. Dubin-Thaler, Olivier Rossier, Yunfei Cai, Oleg Chaga, GuoyingJiang, William Beaver, Hans-Gunther Dobereiner, Yoav Freund, Gary Borisy and Michael P.Sheetz Lamellipodial Actin Mechanically Links Myosin Activity with Adhesion-Site FormationCell Vol 128, 561-575, 09 February 2007

German Creamer and Yoav Freund. A Boosting Approach for Automated Trading. AlgorithmicTrading III Spring 2007, pp. 6778.

Kharchenko P, Chen L, Freund Y, Vitkup D, Church GM. Identifying metabolic enzymes with multipletypes of association evidence. BMC Bioinformatics March 2006

Kundaje A, Middendorf M, Shah M, Wiggins CH, Freund Y, Leslie C. A classification-based frame-work for predicting and analyzing gene regulatory response. BMC Bioinformatics 2006, 7(SupplI):S5

Rui Kuang, Eugne Ie, Ke Wang, Mahira Siddiqi, Yoav Freund and Christina Leslie. Profile-basedstring kernels for remote homology detection and motif extraction. Journal of bio-informaticsand computational biology 2005 June; 3 (3):527-50

M. Middendorf, A. Kundaje, M. Shah, Y. Freund, C. Wiggins and C. Leslie. Motif discovery throughpredictive modeling of gene regulation. RECOMB 2005

3

German Creamer and Yoav Freund. Predicting Performance and Quantifying Corporate GovernanceRisk for Latin American ADRs and Banks. Proceedings of the 2nd IASTED International Con-ference on Financial Engineering and Applications 2004.

M. Middendorf, A. Kundaje, C. Wiggins, Y. Freund and C. Leslie. RECOMB Satellite Workshop onRegulatory Genomics, 2004.

Manuel Middendorf, Anshul Kundaje, Chris Wiggins, Yoav Freund and Christina Leslie. Predictinggenetic regulatory response using classification. Proceedings of the Twelfth International Confer-ence on Intelligent Systems in Molecular Biology (ISMB 2004) 2004.

Yoav Freund, Yishay Mansour and Robert E. Schapire. Generalization bounds for averaged classifiers.The Annals of Statistics Vol. 32, No. 4, 1698–1722, August 2004. A preliminary version appearedin the Proceedings of the Eighth International Workshop on Artificial Intelligence and Statistics,2001.

Yoav Freund and Manfred Opper. Drifting games and Brownian motion. Journal of Computer andSystem Sciences, 64:113–132, 2002. A preliminary version appeared in the Proceedings of the13th Annual Conference on Computational Learning Theory, 2000.

Anat Levin, Paul Viola and Yoav Freund. Unsupervised Improvement of Visual Detectors using Co-Training. Proceedings of the International Conference on Computer Vision (ICCV), Oct 2003,Nice, France.

Yoav Freund and Robert E. Schapire. Discussion of the paper “Additive logistic regression: a sta-tistical view of boosting” by J. Friedman, T. Hastie and R. Tibshirani. The Annals of Statistics,38(2):391–393, April, 2000.

Yoav Freund. An adaptive version of the boost by majority algorithm. Machine Learning, 43(3):293–318, June 2001. A preliminary version appeared in the Proceedings of the 12th Annual Conferenceon Computational Learning Theory, 1999.

Yoav Freund and Robert E. Schapire. A short introduction to boosting. Journal of Japanese Societyfor Artificial Intelligence, 14(5):771–780, September, 1999. (Appearing in Japanese, translationby Naoki Abe.)

Yoav Freund and Yishay Mansour. Estimating a mixture of two product distributions. In Proceedingsof the 12th Annual Conference on Computational Learning Theory, 1999.

Yoav Freund and Llew Mason. The alternating decision tree learning algorithm. In Proceedings ofthe Sixteenth International Conference on Machine Learning, pages 124–133, 1999.

Yoav Freund and Robert E. Schapire. Large margin classification using the perceptron algorithm.Machine Learning, 37(3):277–296, 1999. A preliminary version appeared in the Proceedings ofthe 11th Annual Conference on Computational Learning Theory, 1998.

Yoav Freund, Raj Iyer, Robert E. Schapire and Yoram Singer. An efficient boosting algorithm forcombining preferences. Journal of Machine Learning Research 4:933-969, 2003. A preliminaryversion appeared in the Proceedings of the Fifteenth International Conference on Machine Learn-ing, 1998.

Yoav Freund. Self bounding learning algorithms. In Proceedings of the 11th Annual Conference onComputational Learning Theory, 1998.

Yoav Freund and Robert E. Schapire. Discussion of the paper “Arcing classifiers” by L. Breiman. TheAnnals of Statistics, 26(3):824–832, 1998.

4

Yoav Freund. Improving classification by boosting. In Proceedings of the AT&T conference on Quan-titative Analysis, 1998.

Robert E. Schapire, Yoav Freund, Peter Bartlett and Wee Sun Lee. Boosting the margin: A newexplanation for the effectiveness of voting methods. The Annals of Statistics, 26(5):1651–1686,1998. A preliminary version appeared in the Proceedings of the 14th International Conference onMachine Learning, 1997.

Yoav Freund, Robert E. Schapire, Yoram Singer and Manfred K. Warmuth. Using and combiningpredictors that specialize. In Proceedings of the Twenty-Ninth Annual ACM Symposium on theTheory of Computing, pages 334–343, 1997.

Yoav Freund, H. Sebastian Seung, Eli Shamir, and Naftali Tishby. Selective sampling using the queryby committee algorithm. Machine Learning, 28(2-3):133–168, 1997.

Yoav Freund and Yishay Mansour. Learning under persistent drift. In Proceedings of the third Euro-pean Conference on Computational Learning Theory, pages 109–118, 1997.

Yoav Freund and Robert E. Schapire. Experiments with a new boosting algorithm. In the Proceedingsof the Thirteenth International Conference on Machine Learning, pages 148–156, 1996.

Yoav Freund and Robert E. Schapire. Adaptive game playing using multiplicative weights. Gamesand Economic Behavior, 29:79–103, 1999. A preliminary version appeared in the Proceedings ofthe Ninth Annual Conference on Computational Learning Theory, pages 325–332, 1996.

Yoav Freund. Predicting a binary sequence almost as well as the optimal biased coin. Informationand Computation, 182:73–94, 2003. A preliminary version appeared in the Proceedings of theNinth Annual Conference on Computational Learning Theory, 1996.

Yoav Freund and Robert E. Schapire. A decision-theoretic generalization of on-line learning andan application to boosting. Journal of Computer and System Sciences, 55(1):119–139, 1997. Apreliminary version appeared in the Proceedings of the Second European Conference on Compu-tational Learning Theory, 1995.

Peter Auer, Nicolo Cesa-Bianchi, Yoav Freund, and Robert E. Schapire. The non-stochastic multi-armed bandit problem. SIAM Journal on Computing, 32(1):48-77, 2002. A preliminary versionappeared in the Proceedings of the 36th Annual Symposium on Foundations of Computer Science,1995.

Meir Feder, Yoav Freund, and Yishay Mansour. Optimal universal learning and prediction of proba-bilistic concepts. In Proceedings of the International Symposium on Information Theory, 1995.

Yoav Freund and Dana Ron. Learning to model sequences generated by switching distributions. InProceedings of the Eighth Annual Conference on Computational Learning Theory, pages 41–50,1995.

Yoav Freund, Michael Kearns, Yishay Mansour, Dana Ron, Ronitt Rubinfeld, and Robert E. Schapire.Efficient algorithms for learning to play repeated games against computationally bounded adver-saries. In 36th Annual Symposium on Foundations of Computer Science, pages 332–341, 1995.

Yoav Freund, Michael Kearns, Dana Ron, Ronitt Rubinfeld, Robert E. Schapire, and Linda Sellie.Efficient learning of typical finite automata from random walks. Information and Computation,138(1):23–48, 1997. A preliminary version appeared in the Proceedings of the Twenty-Fifth An-nual ACM Symposium on the Theory of Computing, 1993.

5

Nicolo Cesa-Bianchi, Yoav Freund, David P. Helmbold, David Haussler, Robert E. Schapire, andManfred K. Warmuth. How to use expert advice. Journal of the Association for Computing Ma-chinery, 44(3):427–485, 1997. A preliminary version appeared in the Proceedings of the Twenty-Fifth Annual ACM Symposium on the Theory of Computing, 1993.

Nicolo Cesa-Bianchi, Yoav Freund, David P. Helmbold, and Manfred K. Warmuth. On-line predictionand conversion strategies. Machine Learning, 25:71–110, 1996. A preliminary version appearedin the Proceedings of the first European Conference on Computational Learning Theory, 1993.

Yoav Freund and David Haussler. Unsupervised learning of distributions of binary vectors using 2-layer networks. In Advances in Neural Information Processing Systems, volume 4, pages 912–919,1992.

Yoav Freund. Boosting a weak learning algorithm by majority. Information and Computation,121(2):256–285, 1995. Preliminary versions appeared in the Proceedings of the Third AnnualWorkshop on Computational Learning Theory, 1990, and in the Proceedings of the Fifth AnnualACM Workshop on Computational Learning Theory, 1992.

PatentsYoav Freund and Robert E. Schapire. Apparatus and methods for machine learning hypotheses. US

Patent #5819247, 1998.

Yoav Freund. Internet access system and method with active link status indicators. US Patent#5870769, 1999.

Yoav Freund. Alternating Tree Based Classifiers And Methods For Learning Them. US Patent#6456993, 2002.

6

Grants - recievedYoram Singer, Robert Schapire and Yoav Freund. Theory and applications of boosting algorithms

and margin classifiers. US-Israel Binational Science Foundation, 2000-2003.

Self-Adapting Large-scale solver Architecture Principle Investigators: Victor Eijkhout (U. of Ten-nessee), Yoav Freund (Columbia U.), William Gropp (Argonne National Lab), David Keyes(Columbia U.). Granting agency: NSF. Overall funding: $500,000 from 9/1/2004 to 8/31/2007.

SGER: Computational basis for cognition: Concept learning for Vision and Planning Principle In-vestigators: David Waltz, Yoav Freund (Columbia), Eric Baum. Granting agency:NSF. Overallfunding: $100,000 from 8/1/2004 to 7/31/2005.

IIS: Learning from data of low intrinsic dimension Principle Investigators: Yoav Freund, Sanjoy Das-gupta. Granting Agency:NSF. Overall funding: $450,000 from 8/1/2008 to 7/31/2010.

Selected Invited TalksDrifting Games, Boosting and Online Learning. International conference in Machine Learning June

15, 2009. Video of talk available at http://videolectures.net/icml09 freund dgb/

Machine Learning in System Biology September 13-14 2008, Brussels, Belgium.Video of talk available at http://seed.ucsd.edu/∼yfreund/BioDatabases/

European conference on Machine Learning September 15-19, 2008, Antwerp, Belgium.

Mathematical Foundations of Learning Theory June 18-23, 2004, Barcelona, Spain.

Eurandom Workshop on Statistical Learning in Classification and Model Selection, January 15-182003, Eindhoven, The Netherlands.

Professional ServiceOrganized the symposium Predicting more, Assuming less. New approaches to statistical inference

which was part of the 2000 annual meeting of the American Association for the Advancementof Science. Speakers were: Dr. Yoav Freund, Professor Tom Cover, Professor David Haussler,Professor Rakesh Vohra.

Organized the workshop Online Decision Algorithms that took place in DIMACS, Rutgers Uni-versity from July 12 to July 15, 1999. The workshop was organized together with ProfessorRakesh Vohra and brought together some of the world leading researchers in Computer Science,Game Theory, Mathematical Economics, Information Theory and Statistics. For more details seewww.cs.columbia.edu/∼freund/DIMACS.workshop/index.html

Chaired the COLT steering committee from 1997 to 2001.Established the COLT web site: www.learningtheory.orgUnited COLT and EuroCOLT.

Conference chair, Conference on Computational Learning Theory, 1997 (together with Rob Schapire.)

Editorial Board, Journal of machine learning. 1995-2000Member of the Editorial Board for the Journal of Theoretical Computer Science (TCS-A). (2005-

2006)Action editor in the Journal of Machine Learning (Since 2006)

7

Served on 1-2 NSF panels each year since 2004.

Served on the Microscopy Imaging study section for the NIH.

8

ALON ORLITSKY

ECE and CSE Departments, University of California, San Diego

alon@ucsd.edu (858) 822-0228 http://alon.ucsd.edu

Professional Preparation

1982 - 1986 Ph.D., Electrical Engineering, Stanford University, Stanford, CA1981 - 1982 M.Sc., Electrical Engineering, Stanford University, Stanford, CA1977 - 1981 B.Sc., Electrical Engineering, Ben-Gurion University, Israel1977 - 1980 B.Sc., Mathematics, Ben-Gurion University, Israel

Appointments

2011 - 2014 Director, Center for Wireless Communication2007 - present Qualcomm Professor, University of California, San Diego2006 - present Director, Information Theory and Applications Center1997 - present Professor, ECE & CSE Departments, University of California, San Diego1996 - 1997 Quantitative Analyst, D. E. Shaw and Company1986 - 1996 Member of Technical Staff, AT&T Bell Laboratories1992 - 1993 Adjunct Associate Professor, Columbia University

Most-Related Publications

1. J. Acharya, H. Das, A. Orlitsky, and A.T. Suresh. A Unified Maximum Likelihood Approachfor Estimating Symmetric Properties of Discrete Distributions. In ICML, 2017. (Best paperhonorable mention award: 1 best paper and 3 honorable mentions out of 1,701 submissions.)

2. A. Orlitsky, A.T. Suresh, and Y. Wu. Optimal prediction of the number of unseen species.In PNAS, Vol. 113, No. 47, Nov. 22, 2016.

3. M. Falahatgar, M.I. Ohannessian, and A. Orlitsky. Near-Optimal Smoothing of StructuredConditional Probability Matrices In NIPS, 2016.

4. A. Orlitsky and A.T. Suresh. Competitive distribution estimation: Why is Good-Turinggood. In NIPS, 2015. (Outstanding paper award: 2 papers out of 1,838 submissions.)

5. S. Kamath, A. Orlitsky, V. Pichapati, and A.T. Suresh. On learning distributions from theirsamples. In JMLR-COLT, pages 764–796, 2015.

Other Publications

1. M. Falahatgar, A. Orlitsky, V. Pichapati, and A.T. Suresh. Maximum Selection and Rankingunder Noisy Comparisons. In ICML, 2017.

2. J. Acharya, A. Orlitsky, A.T. Suresh, and H. Tyagi. The complexity of estimating Renyientropy. In SODA, pages 1855–1869, 2015.

3. M. Falahatgar, A. Jafarpour, A. Orlitsky, V. Pichapati, and A.T. Suresh. Universal com-pression of power-law distributions. In ISIT, pages 2001–2005, 2015.

4. M. Falahatgar, A. Jafarpour, A. Orlitsky, V. Pichapati, and A.T. Suresh. Faster algorithmsfor testing under conditional sampling. In JMLR-COLT, pages 607-636, 2015.

5. A. T. Suresh, A. Orlitsky, J. Acharya, and A. Jafarpour. Near-optimal-sample estimatorsfor spherical gaussian mixtures. In NIPS, pages 1395–1403, 2014.

Synergistic Activities

Paper Awards: ICML 2017 Best Paper Honorable Mention (1 best paper, 3 HM’s of 1,701 submissions)

NIPS 2015 Outstanding Paper Award (2 papers out of 1,838 submissions)

Co-author and advisor of Student-Paper-Award Recipient, ISIT 2010

IEEE Transactions on Information Theory Paper Award 2006

Co-author and advisor of Student-Paper-Award Recipient, DCC 2003

IEEE W.R.G. Baker Paper Award, 1992

Plenary Talks: STOC 2017 (one of 11 “best theory” talks), IZS 2014, WITMSE 2010

ISIT 2008, The Learning Workshop 2008, ITW 2007, UAI 2004

IT Society: 2nd VP, 2014; 1st VP, 2015; President, 2016; Junior Past President, 2017

ITA Center: Director, ITA Workshop co-organizer, 2006–present

Mentoring: As ITA director helped hire and co-mentor 16 ITA postdocs

4 of the postdocs are women, 11 out of 15 who finished are now faculty

IT Society Mentoring IT Program, 2008-2013

Mentored 5 ITA undergraduate webmasters, including one now corporate VP

SCOTT KLEMMER d.ucsd.edu/srk @DesignAtLarge 10/1/2018

FACULTY APPOINTMENTS2017– UC San Diego · Professor, Cognitive Science and Computer Science & Engineering

2013–2017 UC San Diego · Associate Professor, Cognitive Science and Computer Science & Engineering2013-15 Stanford University · Visiting Associate Professor, Computer Science2011-13 Stanford University · Associate Professor & Bredt Faculty Scholar Development Chair, Computer Science2004-11 Stanford University · Assistant Professor, Computer Science

DEGREES2004 University of California, Berkeley · PhD in Computer Science2001 University of California, Berkeley · MS in Computer Science1999 Brown University · Dual BA w/honors in Art-Semiotics & Computer Science

Graphic Design coursework at the Rhode Island School of Design

HONORS2018 ACM CHI Honorable Mention Paper (CritiqueKit)2017 ACM CHI Honorable Mention Paper (MyriadHub)2015 Harvard GSE Dean's Distinguished Visiting Fellow2014 Pervasive Health, Honorable Mention (RapidRead)2013 ACM Senior Member2013 ACM CHI Best Paper (Webzeitgeist)2011 Stanford Engineering, Bredt Faculty Scholar Development Chair2011 Katayanagi Emerging Leadership Prize2011 ACM CHI Best Paper (Bricolage)2010 ACM CHI Honorable Mention Paper (Blueprint, HelpMeOut)2009 ACM CHI Honorable Mention Paper (Opportunistic)2006 ACM UIST Best Student Paper (Juxtapose)2008 NSF CAREER2008 Sloan Fellowship2007 ACM CHI Best Paper (Exemplar)2006 ACM UIST Best Paper (d.tools)2006 Microsoft Research New Faculty Fellow

LEADERSHIP2015- Learning at Scale Conference Committee Planning Chair2014- UC San Diego Design Lab Co-Founder & Co-Director

online teaching2015- Interaction Design Specialization, Coursera w/Elizabeth Gerber, Jacob Wobbrock2015- Design for Non-Designers, mooc.house w/Andrea Anderson, VP; Sam Yen, CDO, SAP

2012-2015 HCI Online Class, Coursera More than 200k learners enrolled

program (co-)chair2018 ACM L@S: Learning at Scale2011 ACM UIST: Symposium on User Interface Software and Technologies2010 Stanford Computer Forum Annual Meeting2010 Systems Area, ACM CHI: Human Factors in Computing Systems2009 HCIC: Human Computer Interaction Consortium

editorial board2012–18 Human-Computer Interaction2011–18 ACM TOCHI: Transactions on Computer-Human Interaction

advisory board2015- UC San Diego Extension Design Certificate

2014-17 Apropose, Inc. $1.9M in seed funding2014 TU Delft Industrial Design Engineering

2014- CalIT2 Gallery2014 Coursera Data Policy Advisory Committee2013 UNC Charlotte Design + Computation Program

2011-13 Red Giant2011-12 Caltrain Citizens Advisory Committee

2010- California College of the Arts · Interaction Design Program2010 NTT DoCoMo Labs

(co-)editor2014 Transactions on Computer-Human Interaction Special Issue: Online Learning at Scale2014 Human-Computer Interaction Special Issue on Design Thinking2008 IEEE CG&A Special Issue on Mobile Graphics

workshop/panel (co-)chair2014 ACM CHI · Learning Innovations at Scale2013 National Academy of Engineering · Designing and Analyzing Societal Networks - Frontiers of Engineering2006 ACM CSCW · Collaborating Over Paper and Digital Documents2005 Stanford Computer Forum · Mobile Interaction2005 ACM CHI · The Future of User Interface Design Tools2004 IEEEPervasive · Toolkit Support for Interaction in the Physical World

MAJOR PRESENTATIONS (EXCLUDING COLLOQUIA)2016 TedX, San Diego, CA2015 COIL Fischer Speaker, Penn State, PA2015 SXSW, Austin, TX2014 Google I/O 'Moonshot', Designing a World that Teaches Itself, San Francisco, CA2014 Learning Analytics & Knowledge Keynote, Design at Large, Indianapolis, IN2013 Visual Languages and Human-Centric Computing Keynote, Design at Large: The Power of Examples, San Jose,

CA2013 International Computing Education Research Keynote, San Diego, CA2010 National Academy of Engineering · Future of Engineering Education, Irvine, CA

2005, 2010 PARC Forum, Palo Alto, CA2010 First China Symposium on Human-Computer Interaction, Beijing, China2009 Stanford Leading Matters, Los Angeles, CA2008 Cadence Research Keynote, Berkeley, CA2006 Mobile & Ubiquitous Media Keynote, Stanford, CA2005 IBM New Paradigms in Using Computers Plenary, Almaden, CA2005 Information Processing Society Keynote, Tokyo, Japan

· PHD STUDENTS & POST-DOCSVineet PandeyAilie FraserTricia Ngoon

2016 Catherine Hicks (Post-Doc), Learning Researcher, Google2015 Chinmay Kulkarni (Co-Advised w/Michael Bernstein), Assistant Prof., CMU (Siebel Scholar)2013 Ranjitha Kumar, Assistant Prof., UIUC & Apropose (Google PhD Fellow)2013 Jesse Cirimele, Stanford Post-Doc2013 Nicolas Kokkalis, Stanford Post-Doc2011 Neil Patel (Co-Advised w/Tapan Parikh), Awaaz.De2011 Steven Dow (Post-Doc), Assistant Prof., UC San Diego2010 Joel Brandt, Adobe Labs (Stanford Graduate Fellow)2009 Bjoern Hartmann, Associate Prof., UC Berkeley (Stanford Graduate Fellow)

2007 Ron Yeh, Entrepeneur2007 Brian Lee, Palantir

· SERVICEuc san diegoFaculty Search Chair (Design 16-17, CogSci Teaching 15-17, CogSci 15-16), Masters Committee (CSE 16-17),Undergraduate Curriculum (Cogsci 15-16), Faculty Search (CogSci 14-15), PhD Admissions(CSE 2014–;CogSci 2015), Teaching & Learning Space Committee(2014), Advisor to Design UCSD student group(2014–),

program committeeLearning @ Scale (2015,2014), CHI (2009, 2007, 2006), UIST (2008, 2006, 2005, 2004, 2003), HotMobile (2008),ACM Creativity & Cognition (2007), Graphics Interface (2006, 2005), Ubicomp Workshops (2005, 2004)

refereeTOCHI (2011, 2009, 2007, 2003), SIGGRAPH  (2011, 2008, 2006, 2005, 2003, 2001), New Media & Society (2011),UIST  (2016, 2014, 2012, 2010, 2009, 2007, 2002), CACM  (2010), HCI Journal  (2009), CHI  (annually), Ubicomp (2007,2006, 2004, 2003), AIEDAM  (2007), Graphics Interface  (2007), HICCS  (2007), Int. J. of Human-Computer Studies  (2006), CSCW  (2014, 2008, 2006, 2004, 2000), ISEA  (2006), JCSCW  (2005), ICMI  (2005), IEEE Pervasive (2005), Interacting with Computers (2004), IEEE SMC  (2002)

stanfordUndergraduate Research Advisory Panel (2012), CS Executive Committee(2011-12), CS Faculty Hiring(2011-12), Medicine X Advisory Board(2011–), NSF POMI Steering Group(2010-2013), Gates BuildingImprovements (2010–2012), HCI Breadth Czar (2010-2013), e-Day  (2006, 2010, 2011), Symbolic SystemsSteering Committee (2007–2012), Course Assignments (2007-2009), HCI Comprehensive Exam (2006-2009),PhD Admissions (2005-2009), PhD Program Revision Committee (2009), Computer Forum Committee (2009),Faculty Search Committee (2008), Computer Science Strategic Plan (2005), Undergraduate Research (2004)

· PRIMARY PUBLISHED OR CREATIVE WORKGoogle H-Index: 37. In HCI, student authors are generally listed first; supervising faculty generally listed last. Thepremier HCI publication venues are the annual ACM CHI, UIST, & CSCW conferences. The premier journals areACM TOCHI & Human-Computer Interaction. Strong domain-specific venues include DIS, PervasiveHealth, &Learning@Scale.

20181 Score-Group Framing Negatively Impacts Peer Evaluations. Celia Durkin, Federico Rossano, Scott Klemmer.

CSCW: ACM Conference on Computer Supported Cooperative Work.2 Docent: Transforming personal intuitions to scientific hypotheses through content learning and process

training. Vineet Pandey, Justine Debelius, Embriette R. Hyde, Tomasz Kosciolek, Rob Knight, Scott Klemmer.ACM Learning at Scale.

3 Juxtapeer: Comparative Peer Review Yields Higher Quality Feedback and Promotes Deeper Reflection.Julia Cambre, Scott Klemmer, Chinmay Kulkarni. CHI: ACM Conference on Human Factors in ComputingSystems.

4 Interactive Guidance Technique for Improving Creative Feedback. Tricia J Ngoon, C Ailie Fraser, Ariel SWeingarten, Mira Dontcheva, Scott Klemmer. CHI: ACM Conference on Human Factors in Computing Systems.honorable mention

20175 MyriadHub: Efficiently Scaling Personalized Email Conversations with Valet Crowdsourcing. Nicolas

Kokkalis, Chengdiao Fan, Johannes Roith, Michael S. Bernstein, Scott Klemmer. CHI: ACM Conference onHuman Factors in Computing Systems. honorable mention

6 Gut Instinct: Creating Scientific Theories with Online Learners. Vineet Pandey, Amnon Amir, JustineDebelius, Embriette R. Hyde, Tomasz Kosciolek, Rob Knight, Scott Klemmer. CHI: ACM Conference on HumanFactors in Computing Systems.

20167 Reimagining Human Research Protections for 21st Century Science. Cinnamon Bloss, Camille Nebeker,

Matthew Bietz, Deborah Bae, Barbara Bigby, Mary Devereaux, James Fowler, Ann Waldo, Nadir Weibel,Kevin Patrick, Scott Klemmer, Lori Melichar. Journal of Medical Internet Research. 18

8 DiscoverySpace: Suggesting Actions in Complex Software. C. Ailie Fraser, Mira Dontcheva, HolgerWinnemoeller, Sheryl Ehrlich, Scott Klemmer. DIS: ACM Conference on Designing Interactive Systems.

9 Framing Feedback: Choosing Review Environment Features that Support High Quality Peer Assessment.Catherine Hicks, Vineet Pandey, C. Ailie Fraser, Scott Klemmer. CHI: ACM Conference on Human Factors inComputing Systems.

201510 Talkabout: Making distance matter with small groups in massive classes. Chinmay Kulkarni, Julia Cambre,

Yasmine Kotturi, Michael S. Bernstein, Scott Klemmer. CSCW: ACM Conference on Computer SupportedCooperative Work.

11 Structure and messaging techniques for online peer learning systems that increase stickiness. YasmineKotturi, Chinmay Kulkarni, Michael Bernstein, Scott Klemmer. ACM Learning at Scale.

12 PeerStudio: Rapid Peer Feedback Emphasizes Revision and Improves Performance. Chinmay Kulkarni,Michael S Bernstein, Scott R Klemmer. ACM Learning at Scale.

201413 Supporting Crisis Response with Dynamic Procedure Aids. Leslie Wu, Jesse Cirimele, Kristen Leach, Larry

Chu, T Kyle Harrison, Stuart Card, Scott R Klemmer. DIS: ACM Conference on Designing Interactive Systems.14 RapidRead: Step-At-A-Glance Crisis Checklists. Jesse Cirimele, Leslie Wu, Kristen Leach, Stuart Card, Larry

Chu, T Kyle Harrison, Scott R Klemmer. 8th International Conference on Pervasive Computing Technologies forHealthcare. honorable mention

15 Scaling Short-answer Grading by Combining Peer Assessment with Algorithmic Scoring. Chinmay Kulkarni,Richard Socher, Michael S. Bernstein, Scott R. Klemmer. ACM Learning at Scale.

201316 Peer and Self Assessment in Massive Online Classes. Chinmay Kulkarni, Koh Pang Wei, Huy Le, Daniel Chia,

Kathryn Papadopoulos, Justin Cheng, Daphne Koller, Scott R Klemmer. ACM Transactions on Computer-Human Interaction.

17 TaskGenies: Automatically Providing Action Plans Helps People Complete Tasks. Nicolas Kokkalis, ThomasKoehn, Johanes Huebner, Moontae Lee, Florian Schulze, Scott R Klemmer. ACM Transactions on Computer-Human Interaction. 20

18 Power to the Peers: Authority of Source Effects for a Voice-Based Agricultural Information Service in RuralIndia. Neil Patel, Krishna Savani, Paresh Dave, Kapil Shah, Scott R Klemmer, Tapan S Parikh. InformationTechnologies & International Development.

19 Webzeitgeist: Design Mining the Web. Ranjitha Kumar, Arvind Satyanarayan, Cesar Torres, Maxine Lim,Salman Ahmad, Scott R Klemmer, Jerry O Talton. CHI: ACM Conference on Human Factors in ComputingSystems. best paper

20 EmailValet: Managing Email Overload through Private, Accountable Crowdsourcing. Nicolas Kokkalis,Thomas Köhn, Carl Pfeiffer, Dima Chornyi, Michael S. Bernstein, Scott R. Klemmer. CSCW: ACM Conferenceon Computer-Supported Cooperative Work.

201221 Early and Repeated Exposure to Examples Improves Creative Work. Chinmay Kulkarni, Steven P Dow, Scott

R Klemmer. Cognitive Science.22 Shepherding the Crowd Yields Better Work. Steven P. Dow, Anand Kulkarni, Scott R. Klemmer, Bjoern

Hartmann. CSCW: ACM Conference on Computer Supported Cooperative Work.23 Power to the Peers: Authority of Source Effects for a Voice-based Agricultural Information Service in Rural

India. Neil Patel, Krishna Savani, Paresh Dave, Kapil Shah, Scott R. Klemmer, Tapan S. Parikh. ICTD:International Conference on Information and Communication Technologies and Development.

201124 d.tour: Style-based Exploration of Design Example Galleries. Daniel Ritchie, Ankita Arvind Kejriwal, Scott R

Klemmer. UIST: ACM Symposium on User Interface Software and Technology.25 Flexible Tree Matching. Ranjitha Kumar, Jerry O. Talton, Salman Ahmad, Tim Roughgarden and Scott R.

Klemmer. IJCAI: International Joint Conference on Artificial Intelligence.26 Prototyping Dynamics: Sharing Multiple Designs Improves Exploration, Group Rapport, and Results. Steven

P Dow, Julie Fortuna, Dan Schwartz, Beth Altringer, Daniel L Schwartz, and Scott R Klemmer. CHI: ACMConference on Human Factors in Computing Systems.

27 Bricolage: Example-Based Retargeting for Web Design. Ranjitha Kumar, Jerry O. Talton, Salman Ahmad, andScott R. Klemmer. CHI: ACM Conference on Human Factors in Computing Systems. best paper

201028 Parallel Prototyping Leads to Better Design Results, More Divergence, and Increased Self-Efficacy. Steven P

Dow, Alana Glassco, Jonathan Kass, Melissa Schwarz, Daniel Schwartz, Scott R Klemmer. ACM Transactions onComputer-Human Interaction. 17

29 Example-Centric Programming: Integrating Web Search into the Development Environment. Joel Brandt,Mira Dontcheva, Marcos Weskamp, Scott R. Klemmer. CHI: ACM Conference on Human Factors in ComputingSystems. honorable mention

30 What Would Other Programmers Do? Suggesting Solutions to Error Messages. Bjoern Hartmann, DanielMacDougall, Joel Brandt, Scott R. Klemmer. CHI: ACM Conference on Human Factors in Computing Systems.honorable mention

31 d.note: Revising User Interfaces Through Change Tracking, Annotations, and Alternatives. Bjoern Hartmann,Sean Follmer, Antonio Ricciardi, Timothy Cardenas, Scott R. Klemmer. CHI: ACM Conference on HumanFactors in Computing Systems.

32 Designing with Interactive Example Galleries. Brian Lee, Savil Srivastava, Ranjitha Kumar, Ronen Brafman,Scott R Klemmer. CHI: ACM Conference on Human Factors in Computing Systems.

200933 The Efficacy of Prototyping Under Time Constraints. Steven P. Dow, Kate Heddleston, Scott R. Klemmer.

Creativity & Cognition.34 Two Studies of Opportunistic Programming: Interleaving Web Foraging, Learning, and Writing Code. Joel

Brandt, Philip J. Guo, Joel Lewenstein, Mira Dontcheva, and Scott R. Klemmer. CHI: ACM Conference onHuman Factors in Computing Systems. honorable mention

35 Coordinating Tasks on the Commons: Designing for Personal Goals, Expertise, and Serendipity. MichelKrieger, Emily Margarete Stark, Scott R Klemmer. CHI: ACM Conference on Human Factors in ComputingSystems.

36 Toolkit Support for Integrating Physical and Digital Interactions. Scott R. Klemmer, and James A. Landay. HCIJournal. 24

200837 Range: Exploring Implicit Interaction through Electronic Whiteboard Design. Wendy Ju, Brian Lee, and Scott

R Klemmer. CSCW: ACM Conference on Computer-Supported Cooperative Work.38 Iterative Design and Evaluation of an Event Architecture for Pen-and-Paper Interfaces. Ron B. Yeh, Andreas

Paepcke, and Scott Klemmer. UIST: ACM Symposium on User Interface Software and Technology.39 Design As Exploration: Creating Interface Alternatives through Parallel Authoring and Runtime Tuning.

Björn Hartmann, Loren Yu, Abel Allison, Yeonsoo Yang, Scott R. Klemmer. UIST: ACM Symposium on UserInterface Software and Technology. best student paper

40 Integrating Physical and Digital Interactions on Walls. Scott R. Klemmer, Katherine M. Everitt, James A.Landay. HCI Journal. 23

41 Exiting the cleanroom: on ecological validity and ubiquitous computing. Scott Carter, Jennifer Mankoff,Scott R. Klemmer, and Tara Matthews. HCI Journal. 23

200742 Programming by a Sample: Rapidly Creating Web Applications with d.mix. Björn Hartmann, Leslie Wu, Kevin

Collins, Scott R. Klemmer. UIST: ACM Symposium on User Interface Software and Technology.43 Patterns of Collaboration in Design Courses: Team dynamics affect technology appropriation, artifact

creation, and course performance. Heidy Maldonado, Brian Lee, Scott R Klemmer, Roy D Pea. CSCL:Conference on Computer Supported Collaborative Learning.

44 Authoring Sensor-based Interactions by Demonstration with Direct Manipulation and Pattern Recognition.Björn Hartmann, Leith Abdulla, Manas Mittal, Scott R Klemmer. CHI: ACM Conference on Human Factors inComputing Systems. best paper

200645 Reflective Physical Prototyping through Integrated Design, Test, and Analysis. Björn Hartmann, Scott R.

Klemmer, Michael Bernstein, Leith Abdulla, Brandon Burr, Avi Robinson-Mosher, Jennifer Gee. UIST: ACMSymposium on User Interface Software and Technology. best paper

46 How Bodies Matter: Five Themes for Interaction Design. Scott R Klemmer, Björn Hartmann, Leila Takayama.DIS: ACM Conference on Designing Interactive Systems.

47 ButterflyNet: A Mobile Capture and Access System for Field Biology Research. Ron B. Yeh, Chunyuan Liao,Scott R. Klemmer, François Guimbretière, Brian Lee, Boyko Kakaradov, Jeannie Stamberger, AndreasPaepcke. CHI: ACM Conference on Human Factors in Computing Systems.

48 groupTime: Preference-Based Group Scheduling. Mike Brzozowski, Kendra Carattini, Patrick Mihelich, Scott RKlemmer, Jiang Hu, Andrew Y. Ng. CHI: ACM Conference on Human Factors in Computing Systems.

200449 Papier-Mache: Toolkit Support for Tangible Input. Scott R Klemmer, Jack Li, James Lin, James A Landay. CHI:

ACM Conference on Human Factors in Computing Systems.

200350 Books with Voices: Paper Transcripts as a Tangible Interface to Oral Histories. Scott R Klemmer, Jamey

Graham, Gregory J. Wolff, James A Landay. CHI: ACM Conference on Human Factors in Computing Systems.

51 Two Worlds Apart: Bridging the Gap Between Physical and Virtual Media for Distributed DesignCollaboration. Katherine M. Everitt, Scott R Klemmer, Robert Lee, James A Landay. CHI: ACM Conference onHuman Factors in Computing Systems.

200252 Where Do Web Sites Come From? Capturing and Interacting with Design History. Scott R Klemmer, Michael

Thomsen, Ethan Phelps-Goodman, Robert Lee, James A Landay. CHI: ACM Conference on Human Factors inComputing Systems.

53 Embarking on Spoken-Language NL Interface Design. Anoop K Sinha, Scott R Klemmer, James A Landay.The International Journal of Speech Technology. 5

200154 The Designers' Outpost: A Tangible Interface for Collaborative Web Site Design. Scott R Klemmer, Mark W

Newman, Ryan Farrell, Mark Bilezikjian, James A Landay. UIST: ACM Symposium on User Interface Softwareand Technology.

200055 SUEDE: A Wizard of Oz Prototyping Tool for Speech User Interfaces. Scott R Klemmer, Anoop K Sinha, J.

Chen, James A Landay, Nadeem Aboobaker, Annie Wang. UIST: ACM Symposium on User Interface Softwareand Technology.

SHORT2015

1 Introduction to Online Learning at Scale. Benjamin B. Bederson, Daniel M. Russell, Scott Klemmer. TOCHI:ACM Transactions on Human-Computer Interaction.

20142 Introduction to This Special Issue on Understanding Design Thinking. Scott R Klemmer & John M Carroll.

Human-Computer Interaction.3 State of Design: How Design Education Must Change. Don Norman and Scott Klemmer. LinkedIn.

20114 Skintroducing the Future. Scott R Klemmer. Communications of the ACM.

20095 When is Collaborating with Friends a Good Idea?. Heidy Maldonado, Scott R. Klemmer, Roy D. Pea. CSCL:

Computer Supported Collaborative Learning.6 Opportunistic Programming: Writing Code to Prototype, Ideate, and Discover. Joel Brandt, Philip J. Guo,

Joel Lewenstein, Mira Dontcheva, and Scott R. Klemmer. IEEE Software. 26

20087 Finding Inspiration for the Future in Our Past. Scott R. Klemmer. Ambidextrous Magazine.

8 Hacking, Mashing, Gluing: Understanding Opportunistic Design. Björn Hartmann, Scott Doorley and Scott R.Klemmer. IEEE Pervasive Computing . 7

20059 Teaching Embodied Interaction Design Practice. Scott R. Klemmer, Bill Verplank, Wendy Ju. DUX: ACM

Conference on Designing for User eXperience.10 Integrating Physical and Digital Interactions. Scott R Klemmer. IEEE Computer. 3811 HCI at Stanford University. Terry Winograd and Scott Klemmer. Interactions. 12

200212 Informal PUIs: No Recognition Required. James A Landay, Jason I. Hong, Scott R Klemmer, James Lin, Mark

W Newman. AAAI Spring Symposium: Sketch Understanding Workshop.

CHAPTERS2015

1 Learning Design Wisdom By Augmenting Physical Studio Critique With Online Self-Assessment. ChinmayKulkarni and Scott Klemmer. Reframing Quality Assurance in Creative Disciplines: Evidence from Practice.

20142 Peer and Self Assessment in Massive Online Classes. Chinmay Kulkarni, Koh Pang Wei, Huy Le, Daniel Chia,

Kathryn Papadopoulos, Justin Cheng, Daphne Koller, Scott R Klemmer. Design Thinking Research: BuildingInnovators.

20133 Early and Repeated Exposure to Examples Improves Creative Work. Chinmay Kulkarni, Steven P Dow, Scott

R Klemmer. Design Thinking Research: Building Innovation Eco-Systems.

20124 Prototyping Dynamics: Sharing Multiple Designs Improves Exploration, Group Rapport, and Results. Steven

P. Dow, Julie Fortuna, Dan Schwartz, Beth Altringer, Daniel L. Schwartz, and Scott R. Klemmer. Design ThinkingResearch: Measuring Performance.

20115 Parallel Prototyping Leads to Better Design Results, More Divergence, and Increased Self-Efficacy. Steven P

Dow, Scott R Klemmer. Design Thinking Research: Studying Co-Creation in Practice.

20106 The Efficacy of Prototyping Under Time Constraints. Steven P Dow, Scott R Klemmer. Design Thinking:

Understand - Improve - Apply.7 Programming by a Sample: Leveraging Web Sites to Program Their Underlying Services. Bjoern Hartmann,

Leslie Wu, Kevin Collins, Scott R Klemmer. No Code Required: Giving Users Tools to Transform the Web.8 How the Web Helps People Turn Ideas Into Code. Joel Brandt, Philip J. Guo, Joel Lewenstein, Mira

Dontcheva, Scott R Klemmer. No Code Required: Giving Users Tools to Transform the Web.

20079 Tools for Rapidly Prototyping Mobile Interactions. Yang Li, Scott R Klemmer, James A Landay. Handbook of

Research on User Interface Design and Evaluation for Mobile Technology.

EXTENDED ABSTRACTS

20171 Escaping the Echo Chamber: Ideologically and Geographically Diverse Discussions about Politics. Julia

Cambre, Scott Klemmer, Chinmay Kulkarni. Extended Abstracts of CHI: ACM Conference on Human Factors inComputing Systems.

2 Long-Term Peer Reviewing Effort is Anti-Reciprocal . Yasmine Kotturi, Andrew Du, Chinmay Kulkarni, ScottKlemmer. ACM Learning at Scale: Work in Progress.

3 Standing on the Shoulders of Peers: Tournament-Style Remixing in Project Courses. Julia Cambre, ScottKlemmer. Companion of CSCW: ACM Conference on Computer-Supported Cooperative Work.

20164 Supporting Peer Instruction with Evidence-Based Online Instructional Templates. Tricia Ngoon, Alexander

Gamero-Garrido, Scott Klemmer. ACM Learning at Scale.5 DiscoverySpace: Crowdsourced Suggestions Onboard Novices in Complex Software. C. Ailie Fraser, Mira

Dontcheva, Holger Winnemöller, Scott Klemmer. CSCW: ACM Conference on Computer-SupportedCooperative Work.

20156 Do Numeric Ratings Impact Peer Reviewers?. Catherine M. Hicks, C. Ailie Fraser, Purvi Desai, Scott Klemmer.

ACM Learning at Scale.7 Connecting Stories and Pedagogy Increases Participant Engagement in Discussions. Vineet Pandey, Yasmine

Kotturi, Chinmay Kulkarni, Michael S. Bernstein, Scott Klemmer. ACM Learning at Scale.

20148 Towards Responsive Retargeting of Existing Websites. Gilbert Louis Bernstein and Scott Klemmer. UIST:

ACM Symposium on User Interface Software & Technology.9 Learning Innovation at Scale. Joseph Jay Williams, Rene F. Kizilcec, Daniel M. Russell, Scott R. Klemmer. CHI:

ACM Conference on Human Factors in Computing Systems.10 Community TAs Scale High-Touch Learning, Provide Student-Staff Brokering, and Build Esprit de Corps .

Kathryn Papadopoulos, Lalida Sritanyaratana, and Scott R Klemmer. ACM Learning at Scale.11 Talkabout: Small-group Discussions in Massive Global Classes. Julia Cambre, Chinmay Kulkarni, Michael S.

Bernstein, Scott R. Klemmer. ACM Learning at Scale.

201312 Are MOOCs the Future of Education?. Daniel M. Russell, Scott Klemmer. CHI: ACM Conference on Human

Factors in Computing Systems.13 Tools for Predicting Drop-off in Large Online Classes. Justin Cheng, Chinmay Kulkarni, Scott Klemmer.

CSCW: ACM Conference on Computer-Supported Cooperative Work.

14 Head-mounted and Multi-surface Displays Support Emergency Medical Teams. Leslie Wu, Jesse Cirimele,Jon Bassen, Kristen Leach, Stuart Card, Larry Chu, Kyle Harrison, Scott Klemmer. CSCW: ACM Conference onComputer-Supported Cooperative Work.

201215 Data-Driven Web Design. Ranjitha Kumar, Jerry O. Talton, Salman Ahmad, and Scott R. Klemmer. ICML:

International Conference on Machine Learning.16 A Platform for Large-Scale Machine Learning on Web Design. Arvind Satyanarayan, Maxine Lim, Scott R

Klemmer. CHI: ACM Conference on Human Factors in Computing Systems.

201117 An asymmetric communications platform for knowledge sharing with low-end mobile phones. Neil Patel,

Scott R. Klemmer, Tapan S. Parikh. UIST: Adjunct Proceedings of the ACM symposium on User interface softwareand technology.

18 Maintaining shared mental models in anesthesia crisis care with nurse tablet input and large-screendisplays. Leslie Wu, Jesse Cirimele, Stuart Card, Scott Klemmer, Larry Chu, Kyle Harrison. UIST: AdjunctProceedings of the ACM symposium on User interface software and technology.

19 Shepherding the Crowd: Managing and Providing Feedback to Crowd Workers. Steven P Dow, Brie Bunge,Truc Nguyen, Anand Kulkarni, Bjoern Hartmann, Scott R Klemmer. Works-in-Progress at CHI: ACM Conferenceon Human Factors in Computing Systems.

20 Automatically adapting web pages to heterogeneous devices. Chinmay Kulkarni, Scott R Klemmer. ExtendedAbstracts of CHI: ACM Conference on Human Factors in Computing Systems.

200921 Automatic Retargeting of Web Page Content. Ranjitha Kumar, Juho Kim, and Scott R Klemmer. CHI: work-in-

progress.22 Aesthetics Matter: Leveraging Design Heuristics to Synthesize Visually Satisfying Handheld Interfaces.

Yeonsoo Yang, Scott R. Klemmer. CHI: Work-in-progress.23 Remixing The Web: Enhancing Tailoring Using Programmable Proxies. Joel Brandt, Leslie Wu, Scott R.

Klemmer. CHI Workshop on End User Programming for the Web.

200824 Opportunistic Programming: How Rapid Ideation and Prototyping Occur in Practice. Joel Brandt, Philip J.

Guo, Joel Lewenstein, Scott R. Klemmer. Workshop on End-User Software Engineering IV. ICSE 2008: 30thInternational Conference on Software Engineering.

25 Testing Physical Computing Prototypes Through Time-Shifted & Simulated Input Traces. Timothy Cardenas,Marcello Bastea-Forte, Antonio Ricciardi, Bjoern Hartmann, Scott R. Klemmer. UIST: Extended Abstracts.

26 Rehearse: Coding Interactively while Prototyping. William Choi, Joel Brandt, Scott R. Klemmer. UIST:Extended Abstracts.

200727 Pointer: Multiple Collocated Display Inputs Suggests New Models for Program Design and Debugging.

Marcello Bastea-Forte, Ron Yeh, Scott R. Klemmer. UIST: Extended Abstracts.28 txt 4 l8r: Lowering the Burden for Diary Studies Under Mobile Conditions. Joel Brandt, Noah Weiss, and

Scott R. Klemmer. CHI: Work-in-progress.29 Range: Exploring Proxemics in Collaborative Whiteboard Interaction. Wendy Ju, Brian Lee, and Scott R

Klemmer. CHI: Extended Abstracts. People's Choice Award

200630 Diamond's Edge: From Notebook to Table and Back Again. Michael Bernstein, Avi Robinson-Mosher, Ron

Yeh, Scott Klemmer. Ubicomp: Posters.31 Bridging the Gap: Fluidly Connecting Paper Notecards with Digital Representations for Story/Task-Based

Planning. Tom Hurlbutt, Scott R Klemmer. CHI: Extended Abstracts.32 Lash-Ups: A Toolkit for Location-Aware Mash-Ups . Joel Brandt and Scott R Klemmer. UIST: ACM Symposium

on User Interface Software and Technology.33 Technology for Design Education: A Case Study. Heidy Maldonado, Brian Lee, Scott R Klemmer. CHI:

Extended Abstracts.34 Interaction Design for Active Bodies: Two Themes. Scott R Klemmer, Bjoern Hartmann, Leila Takayama. CHI:

ACM Conference on Human Factors in Computing Systems: Workshop on What is the Next Generation ofHuman-Computer Interaction?.

200535 d.tools: Visually Prototyping Physical UIs through Statecharts. Björn Hartmann, Scott R Klemmer, Michael

Bernstein, Nirav Mehta. UIST: Extended Abstracts.36 ButterflyNet: Mobile Capture and Access for Biologists. Ron B Yeh, Scott R Klemmer. Conference Supplement

to UIST 2005: ACM Symposium on User Interface Software and Technology: Posters.37 Pair Programming: When and Why it Works. Jan Chong, Robert Plummer, Larry Leifer, Scott R Klemmer,

Ozgur Eris, and George Toye. Psychology of Programming Interest Group Workshop.

200338 Papier-Mache: Toolkit Support for Tangible Interaction. Scott R Klemmer. Conference Supplement to UIST

2003: ACM Symposium on User Interface Software and Technology: Doctoral Symposium.

200239 A Pervasive Computing Framework Supporting Collaboration In Documentary History Projects. Scott R

Klemmer. DIS 2002: ACM Conference on Designing Interactive Systems: Post-graduate Symposium.40 Bridging Physical and Electronic Media for Distributed Design Collaboration. Scott R Klemmer, Katherine M.

Everitt. Extended Abstracts of CHI: ACM Conference on Human Factors in Computing Systems.

200141 Supporting Children's Collaboration Across Handheld Computers. Regan Mandryk, Kori M. Inkpen, Mark

Bilezikjian, Scott R Klemmer, James A. Landay. CHI: Work-in-progress.42 SUEDE: Iterative, Informal Prototyping for Speech Interfaces. Anoop K Sinha, Scott R Klemmer, Jack Chen,

James A Landay, Cindy Chen. Extended Abstracts of ChI 2001: ACM Conference on Human Factors inComputing Systems.

43 Different Strokes For Different Folks: A Fluid Toolbelt Of Paper, Walls, And Electronic Sketching. Scott RKlemmer, James A Landay. CHI 2001: ACM Conference on Human Factors in Computing Systems: Workshopon Tools, Conceptual Frameworks, and Empirical Studies for Early Stages of Design.

200044 The Designers' Outpost: A Task-centered Tangible Interface for Web Site Information Design. Scott R

Klemmer, Mark W Newman, Raecine S Sapien. CHI: Conference on Human Factors in Computing Systems.45 A Tangible Difference: Participatory Design Studies Informing a Designers' Outpost. Scott Klemmer, Mark

Newman, Ryan Farrell, Raecine Meza, James A. Landay. CSCW 2000: ACM Conference on ComputerSupported Cooperative Work: Workshop on Shared Environments to Support Face-to-Face Collaboration.

TECH REPORTS2016

1 Scaling Beliefs about Learning to Predict Performance and Enjoyment in an Online Course. Catherine Hicks& Scott Klemmer. PsyArXiv.

2 An HCI View of Configuration Problems. Tianyin Xu, Vineet Pandey, Scott Klemmer. Arxiv.

20123 Our experience with self-assessment and peer critique in design education. Chinmay Kulkarni, Scott R

Klemmer. Tech Report.

20104 Rehearse: Helping Programmers Adapt Examples by Visualizing Execution and Highlighting Related Code.

Joel Brandt, Vignan Pattamatta, Scott R Klemmer. tech-report.

20095 The Effect of Parallel Prototyping on Design Performance, Learning, and Self-Efficacy. Steven P. Dow, Alana

Glassco, Jonathan Kass, Melissa Schwarz, Scott R. Klemmer. Unpublished Technical Report.

20006 Towards a location-based context-aware sensor infrastructure. Scott Klemmer, Sarah Waterson, Kamin

Whitehouse. CS Division, EECS Department, University of California at Berkeley.7 A Tangible Evolution: System Architecture and Participatory Design Studies of the Designers' Outpost. Scott

R Klemmer, Mark W Newman, Ryan Farrell, Raecine Meza, James A Landay. NCSTSTRL.UCB /CSD -00-1116.University of California, Berkeley, Technical Report..

8 Exploring a New Interaction Paradigm for Collaborating on Handheld Computers. Mark Bilezikjian, Regan L.Mandryk, Scott R Klemmer, Kori Inkpen, James A Landay. NCSTSTRL.UCB /CSD -00-1117. University ofCalifornia, Berkeley, Technical Report.

TEACHINGwinter 2016

cogs120/cse170 Interaction Design

fall 2015cogs230/cse216 Interaction Design Research

cogs160 Advanced Interaction Designcogs229/cse219 Design at Large Seminar

winter 2015Cogs120/CSE170 Human-Computer Interaction Design

fall 2014Cogs229/Cse219 Design at LargeCogs230/CSE216 Research in Human-Computer Interaction Design

winter 2014cogs120/cse170 Introduction to Human-Computer Interaction Design

fall 2013cogs230/cse270 Research Topics in Human-Computer Interactioncogs260/cse290 Design at Large Seminar

fall 2012cs147 Intro to Human-Computer Interaction Design 249 studentscs077 Interaction Design Basics 4 students

cs547 Human-Computer Interaction Seminar 95 students

spring 2012cs376 Research Topics in Human-Computer Interaction 32 students

fall 2011cs147L Intro to HCI Labcs147 Introduction to Human-Computer Interaction Design 160 students

spring 2011cs303 Designing Computer Science Experiments 7 studentscs376 Research Topics in Human-Computer Interaction 35 students

fall 2010cs147 Introduction to Human-Computer Interaction Design 150 students

spring 2010cs376 Research Topics in Human-Computer Interaction 31 studentscs303 Designing Computer Science Experiments 9 students

fall 2009cs147 Introduction to Human-Computer Interaction Design 130 students

fall 2008cs147 Introduction to Human-Computer Interaction Design 140 students

cs294h Social Software 14 students

spring 2008cs376 Research Topics in Human-Computer Interaction 28 students

fall 2007cs147 Introduction to Human-Computer Interaction Design 126 studentscs547 Human-Computer Interaction Seminar 81 students

winter 2007cs247 Human-Computer Interaction Design Studio 33 students

fall 2006cs376 Research Topics in Human-Computer Interaction 16 students

spring 2006cs294h Integrating Physical and Digital Interactions 15 students

winter 2006

cs247 Human-Computer Interaction Design Studio 32 students

fall 2005cs376 Research Topics in Human-Computer Interaction 27 students

spring 2005cs377a Mobile Interaction 20 students

winter 2005cs247a Human-Computer Interaction Design Studio 23 students

fall 2004cs376 Research Topics in Human-Computer Interaction 18 students

1

Biographical Sketch

Nadir Weibel

Department of Computer Science and Engineering

University of California, San Diego (UCSD) +1 858 534 8637

9500 Gilman Drive, MC 0404 weibel@ucsd.edu

La Jolla, CA 92093-0404 http://weibel.ucsd.edu

A. PROFESSIONAL PREPARATION

ETH Zurich, Switzerland Computer Science BS 2001, MS 2003, PhD 2009

University of California San Diego Human-Computer Interaction Postdoctoral fellow, 2009-2012

B. ACADEMIC/PROFESSIONAL APPOINTMENTS

07/17 –

Associate Research Professor, Department of Computer Science and Engineering,

UCSD

07/13 – 07-17

Assistant Research Professor, Department of Computer Science and Engineering,

UCSD

01/13 – Research Health Science Specialist, VA San Diego Research Service

09/12 – Lecturer, Computer Science and Engineering & Cognitive Science, UCSD

10/ 09 – 09/12 Postdoctoral Fellow, Department of Cognitive Science, UCSD

09/08 – 08/09 Research engineer and solution consultant, Sungard Financial Systems, Zurich,

Switzerland

09/08 – 12/11 Lecturer and course coordinator, DECS Cantone Ticino, Switzerland

09/04 – 08/11 Research and teaching assistant, ETH Zurich, Switzerland

09/03 – 08/04 Networking consultant and Exams Coordinator, Cantone Ticino, Switzerland

09/03 – 08/04 Software consultant / lecturer, Centro Professionale Trevano, Lugano, Switzerland

C. PUBLICATIONS

Publications Most Closely Related to Proposal

• D. Gasques Rodrigues, A. Jain, S. Rick, P. Suresh, S. Liu, and N. Weibel, “Exploring Mixed Reality in Specialized Surgical Environments,” in Proceedings of CHI 2017 (Late Breaking), ACM Conference on Human Factors in Computing Systems, Denver (CO), USA, 2017.

• N. Weibel, S. Rick, C. Emmenegger, S. Ashfaq, A. Calvitti, and Z. Agha. LAB-IN-A-BOX: Semi-Automatic Tracking of Activity in the Medical Office. Pers Ubiquit Comput - Health, pages 11-18, August 2014

• S. Rick, V. Ramesh, D. Gasques Rodrigues, and N. Weibel, “Pervasive Sensing in Healthcare: From Observing and Collecting to Seeing and Understanding,” in In Proc. of WISH, Workshop on Interactive System for Healthcare, CHI 2017.

• N. Weibel and J.D. Hollan. Gesture and Action Recognition. In Abstracts of ISGS 2014, International Society of Gestures Studies 6, San Diego, USA, July 2014

• A. Fouse, N. Weibel, E. Hutchins, and J. Hollan. ChronoViz: A System for Supporting Navigation of Time-coded Data. In Proc. CHI 2011, pages 299-304

Other Significant Publications

• Vaizman, Y., Weibel, N., and Lanckriet, G. "Context Recognition In-the-Wild: Unified Model for Multi-Modal Sensors and Multi-Label Classification". Proceedings of the ACM on Interactive, Mobile, Wearable and Ubiquitous Technologies (IMWUT), vol. 1, no. 4. December 2017.

2

• Vaizman, Y., Ellis, K., Lanckriet, G., and Weibel, N. "Data Collection In-the-Wild with the ExtraSensory App: Rich User Interface to Self-Report Behavior". Computer Human Interaction (CHI 2018), Montreal, April 2018.

• B. Balaji, J. Koh, N. Weibel, and Y. Agarwal, “Genie: A Longitudinal Study Comparing Physical and Software Thermostats in Office Buildings,” in Proceedings of Ubicomp 2016, ACM International Joint Conference on Pervasive and Ubiquitous Computing, Heidelberg, Germany, 2016.

• E. Tanuwidjaja, D. Huynh, K. Koa, C. Nguyen, C. Shao, P. Torbett, C. Emmenegger, and N. Weibel. Chroma: A Wearable Augmented-Reality Solution for Color-Blindness. In Proc. Ubicomp 2014. In Press

• D. Kusunoki, A. Sarcevic, N. Weibel, I. Marsic, Z. Zhang, G. Tuveson, and R. Burd. Balancing Design Tensions in the Trauma Bay: Iterative Display Design to Support Ad Hoc and Interdisciplinary Medical Teamwork. In Proc. CHI 2014, pages 3777–3786

• A. Piper, N. Weibel, and J. Hollan. Audio-Enhanced Paper Photos: Encouraging Social Interaction at Age 105. In Proc. CSCW 2013, pages 215–224

• E. Hutchins, N. Weibel, C. Emmenegger, A. Fouse, B. Holder. An Integrative Approach to Understanding Flight Crew Activity, Journal of Cognitive Engineering and Decision Making, July 2013

• N. Weibel, A. Fouse, C. Emmenegger, W. Friedman, W. Hutchins, and J. Hollan. Digital Pen and Paper Practices in Observational Research. In Proc. CHI 2012, pages 1331-1340.

D. SYNERGISTIC ACTIVITIES

-Program Committee member: PervasiveHealth 2018 (Program Chair), Ubicomp 2016 (DC Co-

Chair), Ubicomp 2015, CHI 2014 (Assoc. Chair), SenseCam 2013 (Program Chair), CHI 2013+2014

(Assoc. Chair), PervasiveHealth 2013, UCAml 2012, IWAAL 2012, IE2012, IUI 2012.

-Reviewer: CHI, UIST, IUI, PervasiveHealth, EICS, DIS, ITS, CSCW, Ubicomp, Int. J of Hum-

Comp Study, ACM TOCHI, ACM TIIS, Springer PUC, the Swiss State Secretariat for Education

and Research Program, NSF HCC 2013+2018

-Organizing Committee: WISH at CHI 2019, PervasiveHealth 2016 (General Co-Chair), CHI

2013+2014, SenseCam 2013, PervasiveHealth 2012+2013, IUI 2012, CoPADD 2006 +2007.

-Broadening Participations: UCSD’s CSE Department “Women in Technology” (WIT) Committee

-Research Tools and Innovation: CocoonCam, ChronoViz, Lab-in-a-Box, HoloCPR,

HoloUltrasound

Supplementary Information for Remote Instruction Course Proposals (for all Masters of Data Science Courses) Per the UC San Diego Policy on Distance Education Courses, we answer the following 5 questions. These answers apply to all courses in the online Masters of Data Science. 1. How will the course content be delivered (e.g. Learning Management System, online textbook/videos, video hosting platforms, lecture formats, etc.)? Course content will be delivered through a combination of videos, readings, links for supplemental resources, quizzes, exams, and projects.

● Video lectures. Video mini-lectures (4-20 minutes) teach course content as well as provide live-coding examples which encourage students to work through their own code in parallel with the instructor. In addition, there are a series of guest lectures which introduce students to data science faculty at UCSD and SDSC.

● Readings. Students are given brief readings with either code or course content to review.

● Discussion forum prompts. Students are prompted periodically to engage in discussions regarding course content. They are given questions to help facilitate that discussion.

● Formative quizzes. Ungraded short practice quizzes allow students to get feedback on their learning periodically in the course.

● Graded Quizzes. Graded quizzes are short quizzes that are not worth a large portion of the grade. Students will not be proctored during these quizzes. This mirrors a practice for in-person courses where some instructors use online graded quizzes that are not proctored.

● Projects. Some classes will have project components, which will be graded using a variety of approaches, including: (1) automated grading (2) TA/Instructor grading (3) peer grading with careful rubrics and appealable grades (note that peer grading is also a practice that is beginning to gain traction for in-person classes)

● Exams. Exams will be proctored using Software Secure, a leading e-proctoring system that verifies student identities and monitors students during exams

2. What technologies/tools will be used for student-instructor interaction, student-IA interaction, student-student interaction, and instructor-IA interactions? Indicate: a. the frequency of these interactions; b. whether the interactions are required or optional; and c. whether the interactions are asynchronous or synchronous.

As with other courses, we will hire graduate students to be the Instruction Assistants (IAs) for the course. The IAs will monitor forums, answer student questions, assist with peer review when needed, and hold Office Hours.

Office hours: The instructor and each IA will hold two kinds of office hours: scheduled office hours, and ad-hoc office hours.

Scheduled office hours will happen at advertised times. done either through online video chatting (using tools like Google hangouts, Google chat), or through online questions forums (where the idea is that a certain assigned times, there will be someone monitoring questions non-stop and answering right away) The instructor will use feedback throughout the class to fine tune the amount of scheduled office hours, and appropriate times for these office hours (especially considering the global audience of the program, which includes issues with time zones)

The second kind of office hours will be ad-hoc office hours. These will consist of instructors and IAs logging into online forums periodically to answer questions that have been posted. For example, each IA might be asked to spend 30 minutes a day answering questions throughout the day (for example, go online 6 times a day and answer questions for 5 minutes each time).

Instructor role: The instructor will monitor forums to make sure that questions are answered appropriately. In an online setting, the instructor can also weigh in on questions that have already been answered. In fact, office hours through an online forum provide much more oversight than traditional in-person office hours. Indeed, in a traditional setting, the in-person interactions between a IA and students are not recorded and not looked at by the instructor. If a IA answers a question incorrectly (or just not as well as they should have) during in-person office hours, neither the student nor the instructor find out about it. However, in an online forum, the openness and transparency provides oversight not only from the instructor but from other IAs and other students, who can also weigh in on the discussion.

By seeing all the questions that are asked, the instructor can use the online office hours and online forums to get a good sense of where student misconceptions are. The instructor can then address these misconceptions by providing additional content, or additional assignments. Furthermore, in an online setting, there is also the potential for a fine-grained assessment of student mis-understandings. Indeed, because students are progressing in the course online (which can include practice quizzes to test understanding), the instructor can look at the rate of progression of students to pinpoint where students are getting confused. For example, if the instructor notices that, say question 3 of the test quiz of the second online lecture is disproportionately answered wrong, the instructor will know that the concept covered by that question is a source of confusion.

To answer more specifically:

a. the frequency of these interactions: the frequency will depend on the course and the instructor, but will match the frequency of in-person courses.

b. whether the interactions are required or optional: interactions between IAs and instructors will be mandatory; interactions between students and IAs will be optional, in the same way that they are currently optional for in-person courses.

c. whether the interactions are asynchronous or synchronous: it will be a combination of the two.

3. How will students be evaluated (e.g. quizzes, written assignments, problems sets, final exam, final paper, final presentation)? Describe the frequency of the evaluations and the type of feedback students receive. Several approaches will be used to evaluate students:

● Graded and Ungraded Quizzes ● Proctored Exams during the quarter ● Proctored Final Exams ● Graded Assignments ● Graded projects throughout the quarter ● For courses where there is one large project at the end of the quarter, graded milestones

will be used throughout the quarter Each course will use a subset of the above techniques, depending on the material being taught. 4. Describe how student identity will be verified, especially for high stake assessments like midterms and final exams. How will academic integrity be handled? Section 2.3.6 of the proposal for the new Masters of Data Science describes our efforts at ensuring academic integrity, including the appointment of an Academic Integrity Coordinator who will oversee all efforts related to academic integrity for the program. To help ensure integrity during exams, we will use Software Secure, a leading e-proctoring system that verifies student identities and monitors students during exams. This is the standard for online masters programs. When an exam starts, Software Secure authenticates students using a photo ID through a webcam, and requires a room scan of the test environment. During the exam, Software Secure monitors and records the learner (through a webcam) and computer desktop (using desktop monitoring software) to catch any inappropriate physical or online activity. The course team can specify the rules (e.g.: no phones, no calculators, no going to other websites, etc.). Any suspicious event is reported by the Software Secure team to the instructor at UC San Diego to make a determination. Each proctored event (audio + video) is recorded and can later be replayed to look at problematic cases, thus providing a significant level of monitoring. This kind of e-proctoring has already been approved by UCSD Graduate Council for graduate R (remote) classes.

We will also use plagiarism detection software to make sure that students are not copying from each other or from online sources. We will also encourage Academic Integrity in the syllaby of the courses. Here is a sample from the MDS 200 course: Excelling with Integrity

You are in this course because you want to learn and we want to do everything we can to help you learn.

In order for that to happen, you need to do your own work and not help other people do work they

should be doing.

As we tell our in-person students, focusing on course grades and course credit is short-sighted. We take

courses to learn new skills and to learn new ways of thinking about the world. Yes, you get grades for

your work in a course. But grades and course credit just helps open doors for you - with these grades

comes the expectation you know the course material. When it comes time for you to use your

knowledge and skills, you'll want the pride and personal confidence of knowing you did the work

yourself.

We care a lot of about this and so should you. If you're unclear about what is permitted and what is not,

check out the agreement and if still in doubt, just ask.

5. If the course employs IAs, describe how the IAs will interact with the students and provide the student/IA ratio. Describe how the IAs will be trained, and how the IAs will interact with instructors.

Section 5.1 of the proposal for the Master of Data Science covers TA training. The instructor will be responsible for training the IAs to make sure they are effective in the setting of an online class. This will include making them familiar with the online platforms, understanding how to answer questions online, and how to hold online office hours. Even though the class is online, the instructor and the IAs will meet in person once a week to manage the class (and of course will coordinate through email and other online forums more frequently). Training of the IAs is further addressed in Section 5.1 of our proposal.

Supplementary Information for Remote Instruction Course Proposals (for all Masters of Data Science Courses) Per the CEP Policy on Remote and Distance Instruction, we answer the following 5 questions. These answers apply to all courses in the online Masters of Data Science. 1. In the absence of regularly scheduled meetings of students with the instructor in a single classroom, how will course content be delivered to students? Course content will be delivered on the edX platform. The course on edX includes videos, readings, links for supplemental resources, quizzes, exams, and projects.

● Video lectures. Video mini-lectures (4-20 minutes) teach course content as well as provide live-coding examples which encourage students to work through their own code in parallel with the instructor. In addition, there are a series of guest lectures which introduce students to data science faculty at UCSD and SDSC.

● Readings. Students are given brief readings with either code or course content to review.

● Discussion forum prompts. Students are prompted periodically to engage in discussions regarding course content. They are given questions to help facilitate that discussion.

● Formative quizzes. Ungraded short practice quizzes allow student to get feedback on their learning periodically in the course.

● Graded Quizzes. Graded quizzes are short quizzes that are not worth a large portion of the grade. Students will not be proctored during these quizzes. This mirrors a practice for in-person courses where some instructors use online graded quizzes that are not proctored.

● Projects. Some classes will have project components, which will be graded using a variety of approaches, including: (1) automated grading (2) TA/Instructor grading (3) peer grading with careful rubrics and appealable grades (note that peer grading is also a practice that is beginning to gain traction for in-person classes)

● Exams. Exams will be proctored using Software Secure, a leading e-proctoring system that verifies student identities and monitors students during exams

edX will host our courses and provide support for student with disabilities. Their full accessibility policy can be found here: https://www.edx.org/accessibility 2. Students must have the opportunity to interact regularly with their instructor, teaching assistants (if applicable), and other students, e.g. to ask questions and exchange ideas. How will this be achieved?

Students can interact with other students via course forums. The courses will have a number of required forum activities to encourage participation and engagement. The majority of student questions are answered by other students. The remaining questions are answered by a combination of the course instructors and a team of students who are acting in a role similar to that of teaching assistants for in-person courses. 3. Remote instruction creates different expectations for teaching assistants (TAs) compared to conventional courses. If the proposed course will involve TAs, how will they be trained to serve their intended role? As with other courses, we will hire graduate students to be the Teaching Assistants (TAs) for the course. The TAs will monitor forums, answer student questions, assist with peer review when needed, and hold Office Hours.

Office hours: The instructor and each TA will hold two kinds of office hours: scheduled office hours, and ad-hoc office hours.

Scheduled office hours will happen at advertised times. done either through online video chatting (using tools like Google hangouts, Google chat), or through online questions forums (where the idea is that a certain assigned times, there will be someone monitoring questions non-stop and answering right away) The instructor will use feedback throughout the class to fine tune the amount of scheduled office hours, and appropriate times for these office hours (especially considering the global audience of the program, which includes issues with time zones)

The second kind of office hours will be ad-hoc office hours. These will consist of instructors and TAs logging into online forums periodically to answer questions that have been posted. For example, each TA might be asked to spend 30 minutes a day answering questions throughout the day (for example, go online 6 times a day and answer questions for 5 minutes each time).

Training TAs: The instructor will be responsible for training the TAs to make sure they are effective in the setting of an online class. This will include making them familiar with the online platforms, understanding how to answer questions online, and how to hold online office hours. Even through the class is online, the instructor and the TAs will meet in person once a week to manage the class (and of course will coordinate through email and other online forums more frequently)

Instructor role: The instructor will monitor forums to make sure that questions are answered appropriately. In an online setting, the instructor can also weigh in on questions that have already been answered. In fact, office hours through an online forum provide much more oversight than traditional in-person office hours. Indeed, in a traditional setting, the in-person interactions between a TA and students are not recorded and not looked at by the instructor. If a TA answers a question incorrectly (or just not as well as they should have) during in-person

office hours, neither the student nor the instructor find out about it. However, in an online forum, the openness and transparency provides oversight not only from the instructor but from other TAs and other students, who can also weigh in on the discussion.

By seeing all the questions that are asked, the instructor can use the online office hours and online forums to get a good sense of where student misconceptions are. The instructor can then address these misconceptions by providing additional content, or additional assignments. Furthermore, in a online setting, there is also the potential for a fine-grained assessment of student mis-understandings. Indeed, because students are progressing in the course online (which can include practice quizzes to test understanding), the instructor can look at the rate of progression of students to pinpoint where students are getting confused. For example, if the instructor notices that, say question 3 of the test quiz of the second online lecture is disproportionately answered wrong, the instructor will know that the concept covered by that question is a source of confusion.

4. Please explain how students will be able to assess their progress in the course as it proceeds. Several approaches will be used to let students know of their progress during the course:

● Graded and Ungraded Quizzes ● Proctored Exams during the quarter ● Graded Assignments ● Graded projects throughout the quarter ● For courses where there is one large project at the end of the quarter, graded milestones

will be used throughout the quarter Each courses will use a subset of the above techniques, depending on the material being taught. 5. Remote instruction courses raise heightened concerns about academic dishonesty. What graded work will be required of students and what measures will be taken to minimize the opportunity for academic dishonesty in completing this work? For example, if the course employs online exams or quizzes, how will you ensure that the enrolled student is the one taking the test and that unauthorized aids are not used? Alternatively, will proctored, in-person exams be given and what choices will students have regarding where and when they can take them? Section 2.3.6 of the proposal for the new Masters of Data Science describes our efforts at ensuring academic integrity, including the appointment of an Academic Integrity Coordinator who will oversee all efforts related to academic integrity for the program. edX has an Honor Code which details appropriate student behavior. In addition to the standard edX honor code, we will have students agree to a standard academic integrity agreement which appears in many existing CSE courses. To help ensure integrity during exams, we will use

Software Secure, a leading e-proctoring system that verifies student identities and monitors students during exams. When an exam starts, Software Secure authenticates students using a photo ID through a webcam, and requires a room scan of the test environment. During the exam, Software Secure monitors and records the learner (through a webcam) and computer desktop (using desktop monitoring software) to catch any inappropriate physical or online activity. The course team can specify the rules (e.g.: no phones, no calculators, no going to other websites, etc.). Any suspicious event is reported by the Software Secure team to the instructor at UC San Diego to make a determination. Each proctored event (audio + video) is recorded and can later be replayed to look at problematic cases, thus providing a significant level of monitoring. This kind of e-proctoring has already been approved by UCSD Graduate Council for graduate R (remote) classes. We will also use plagiarism detection software to make sure that students are not copying from each other or from online sources. We will also encourage Academic Integrity in the syllaby of the courses. Here is a sample from the MDS 200 course: Excelling with Integrity

You are in this course because you want to learn and we want to do everything we can to help you learn.

In order for that to happen, you need to do your own work and not help other people do work they

should be doing.

As we tell our in-person students, focusing on course grades and course credit is short-sighted. We take

courses to learn new skills and to learn new ways of thinking about the world. Yes, you get grades for

your work in a course. But grades and course credit just helps open doors for you - with these grades

comes the expectation you know the course material. When it comes time for you to use your

knowledge and skills, you'll want the pride and personal confidence of knowing you did the work

yourself.

We care a lot of about this and so should you. Please be sure you read the edX terms of service

agreement. If you're unclear about what is permitted and what is not, check out the agreement and if

still in doubt, just ask.

9/11/2018 E Course Request

https://act.ucsd.edu/qlink/easyquery 1/3

UNIVERSITY OF CALIFORNIA - SAN DIEGOE COURSE REQUESTDate: 09/11/2018

Request Type :Subject Code :Course Number :

New

MDS

200

Department :Course Title :Transcript Title :Effective Term :Extent of Approval :Summer Only :

Description :

Computer Science & Engineering

Python for Data Science

Python for Data Science

SPRING 20

NStudents will learn to find answers within large datasets by using Python tools to import data, explore it, analyze it, learn from it, visualize it, and ultimately generate easily shareable reports. Topics include: basic process of data science; Python and Jupyter notebooks; an applied understanding of how to manipulate and analyze uncurated datasets; basic statistical analysis and machine learning methods; how to effectively visualize results.

Course Justification : Foundation course of the proposed online Masters of Data Science

Request Justification : Foundation course of the proposed online Masters of Data Science

FIAT :

Instructors:Instructor Assignment : Department Chair Assigned

Department Title/Rank First Name Initial Last Name

Crosslisted Courses:Subject Code Course Number Department Course Title Status End Term

Instructional Units/Hours:Unit Type : FIXEDFixed Units : 4.00

Variable Units : Min : 0.00 Max : 0.00 Increment : 0.00

Types of Instruction: Fixed Hours Var Hours Min Var Hours Max Grade Report OtherLE 3.00 0.00 0.00 NDI 1.00 0.00 0.00 YOP 8.00 0.00 0.00 N

Total Hours(Fixed) : 12.00 Total(Var)Min/Max :

Course Repeatability:Number of Times Taken forCredit: 1

Total Unit Credits : 4.00Justification :

Grade Options:Undergraduate :Graduate : Standard(letter/or S/U)

9/11/2018 E Course Request

https://act.ucsd.edu/qlink/easyquery 2/3

In Progress Grading:Does course use in-progressgrading : N

IP Justification :

In Progress Sequence Courses:

Status Subject Code Course No. Course Title Seq No. Final Grade End Term

Final Evaluation:Final Exam Method : P3/O

Keys: C=In Class Final Exam; L=Lab Final; P=Final Paper; P2=Final Presentation; P3=Final Project; O=Other; NL=Nofinal(Lab); NG=No Final(Grad Course);

Description : Remotely proctored final exam

Justification : Final project, in conjunction with a remotely proctored final exam,represents the optimal method for final evaluation for this course

Course Prerequisites:

Status Department Subject Code Course No. Title End Term Prerequisite Seq

Test Prerequisites:

Test Description Low Score High Score

Department Restrictions:Department Approval Req : NDepartment Approval Req Justification:Other Enrollment Requirements : Restricted to students within the MDS major code.

Academic Level Restrictions:

Freshman: N Sophomore : N Junior : N Senior : NLower Div: N Upper Div : N Graduate : N Pharmacy : N Medical : N

Course Corequisites:

Status Department Subject Code Course No. Course Title End Term

Course Duplicates:

Status Department Subject Code Course No. Course Title End TermProposed Computer Science DSE 200 Python for Data ScienceProposed Computer Science DSE 200R Python for Data Science

Other Catalog Information:Other Catalog Information(Optional):Recommended Preporation(Optional) :Material Fee : N

Animal Subjects:Uses Animal Subjects: NProtocol Number :Approval Date :

Human Subjects:

9/11/2018 E Course Request

https://act.ucsd.edu/qlink/easyquery 3/3

Uses Human Subjects: NProtocol Number :Approval Date :

Decisions:

Reviewer Department Decision Userid Date

9/11/2018 E Course Request

https://act.ucsd.edu/qlink/easyquery?qlink.report.id=ECourseRequest&A=10751 1/3

UNIVERSITY OF CALIFORNIA - SAN DIEGOE COURSE REQUESTDate: 09/11/2018

Request Type :Subject Code :Course Number :

New

MDS

210

Department :Course Title :Transcript Title :Effective Term :Extent of Approval :Summer Only :

Description :

Computer Science & Engineering

Probability and Statistics in Data Science using Python

Prob/Stats in DataSci w/Python

SPRING 20

NThis course covers the foundations of probability and statistics needed for Data Science. Students will learn both the mathematical theory, and get hands-on experience of applying this theory to actual data using Jupyter notebooks. Covered topics include: Random variables; Dependence; Correlation; Regression; PCA; Entropy and MDL

Course Justification : Foundation course of the proposed online Masters of Data Science

Request Justification : Foundation course of the proposed online Masters of Data Science

FIAT :

Instructors:Instructor Assignment : Department Chair Assigned

Department Title/Rank First Name Initial Last Name

Crosslisted Courses:Subject Code Course Number Department Course Title Status End Term

Instructional Units/Hours:Unit Type : FIXEDFixed Units : 4.00

Variable Units : Min : 0.00 Max : 0.00 Increment : 0.00

Types of Instruction: Fixed Hours Var Hours Min Var Hours Max Grade Report OtherDI 1.00 0.00 0.00 YLE 3.00 0.00 0.00 NOP 8.00 0.00 0.00 N

Total Hours(Fixed) : 12.00 Total(Var)Min/Max :

Course Repeatability:Number of Times Taken forCredit: 1

Total Unit Credits : 4.00Justification :

Grade Options:Undergraduate :Graduate : Standard(letter/or S/U)

9/11/2018 E Course Request

https://act.ucsd.edu/qlink/easyquery?qlink.report.id=ECourseRequest&A=10751 2/3

In Progress Grading:Does course use in-progressgrading : N

IP Justification :

In Progress Sequence Courses:

Status Subject Code Course No. Course Title Seq No. Final Grade End Term

Final Evaluation:Final Exam Method : P3/O

Keys: C=In Class Final Exam; L=Lab Final; P=Final Paper; P2=Final Presentation; P3=Final Project; O=Other; NL=Nofinal(Lab); NG=No Final(Grad Course);

Description : Remotely proctored exam

Justification : Final project, in conjunction with a remotely proctored final exam,represents the optimal method for final evaluation for this course

Course Prerequisites:

Status Department Subject Code Course No. Title End Term Prerequisite Seq

Test Prerequisites:

Test Description Low Score High Score

Department Restrictions:Department Approval Req : NDepartment Approval Req Justification:Other Enrollment Requirements : Restricted to students within the MDS major code.

Academic Level Restrictions:

Freshman: N Sophomore : N Junior : N Senior : NLower Div: N Upper Div : N Graduate : N Pharmacy : N Medical : N

Course Corequisites:

Status Department Subject Code Course No. Course Title End Term

Course Duplicates:

Status Department Subject Code Course No. Course Title End Term

Other Catalog Information:Other Catalog Information(Optional):Recommended Preporation(Optional) :Material Fee : N

Animal Subjects:Uses Animal Subjects: NProtocol Number :Approval Date :

Human Subjects:Uses Human Subjects: NProtocol Number :Approval Date :

9/11/2018 E Course Request

https://act.ucsd.edu/qlink/easyquery?qlink.report.id=ECourseRequest&A=10751 3/3

Decisions:

Reviewer Department Decision Userid Date

9/11/2018 E Course Request

https://act.ucsd.edu/qlink/easyquery?qlink.report.id=ECourseRequest&A=10753 1/3

UNIVERSITY OF CALIFORNIA - SAN DIEGOE COURSE REQUESTDate: 09/11/2018

Request Type :Subject Code :Course Number :

New

MDS

220

Department :Course Title :Transcript Title :Effective Term :Extent of Approval :Summer Only :

Description :

Computer Science & Engineering

Machine Learning Fundamentals

Machine Learning Fndamntls

SPRING 20

NThis course covers supervised and unsupervised learning algorithms, with applications in Python and Jupyter notebooks. Covered topics include: Classification, regression, and conditional probability estimation; Generative and discriminative models; Linear models and extensions to nonlinearity using kernel methods; Ensemble methods: boosting, bagging, random forests; Representation learning: clustering, dimensionality reduction, autoencoders, deep neural networks.

Course Justification : Foundation course of the proposed online Masters of Data Science

Request Justification : Foundation course of the proposed online Masters of Data Science

FIAT :

Instructors:Instructor Assignment : Department Chair Assigned

Department Title/Rank First Name Initial Last Name

Crosslisted Courses:Subject Code Course Number Department Course Title Status End Term

Instructional Units/Hours:Unit Type : FIXEDFixed Units : 4.00

Variable Units : Min : 0.00 Max : 0.00 Increment : 0.00

Types of Instruction: Fixed Hours Var Hours Min Var Hours Max Grade Report OtherDI 1.00 0.00 0.00 YLE 3.00 0.00 0.00 NOP 8.00 0.00 0.00 N

Total Hours(Fixed) : 12.00 Total(Var)Min/Max :

Course Repeatability:Number of Times Taken forCredit: 1

Total Unit Credits : 4.00Justification :

Grade Options:Undergraduate :

9/11/2018 E Course Request

https://act.ucsd.edu/qlink/easyquery?qlink.report.id=ECourseRequest&A=10753 2/3

Graduate : Standard(letter/or S/U)

In Progress Grading:Does course use in-progressgrading : N

IP Justification :

In Progress Sequence Courses:

Status Subject Code Course No. Course Title Seq No. Final Grade End Term

Final Evaluation:Final Exam Method : P3/O

Keys: C=In Class Final Exam; L=Lab Final; P=Final Paper; P2=Final Presentation; P3=Final Project; O=Other; NL=Nofinal(Lab); NG=No Final(Grad Course);

Description : remotely proctored final exam

Justification : Final project, in conjunction with a remotely proctored final exam,represents the optimal method for final evaluation for this course

Course Prerequisites:

Status Department Subject Code Course No. Title End Term Prerequisite Seq

Test Prerequisites:

Test Description Low Score High Score

Department Restrictions:Department Approval Req : NDepartment Approval Req Justification:Other Enrollment Requirements : Restricted to students within the MDS major code.

Academic Level Restrictions:

Freshman: N Sophomore : N Junior : N Senior : NLower Div: N Upper Div : N Graduate : N Pharmacy : N Medical : N

Course Corequisites:

Status Department Subject Code Course No. Course Title End Term

Course Duplicates:

Status Department Subject Code Course No. Course Title End Term

Other Catalog Information:Other Catalog Information(Optional):Recommended Preporation(Optional) :Material Fee : N

Animal Subjects:Uses Animal Subjects: NProtocol Number :Approval Date :

Human Subjects:Uses Human Subjects: N

9/11/2018 E Course Request

https://act.ucsd.edu/qlink/easyquery?qlink.report.id=ECourseRequest&A=10753 3/3

Protocol Number :Approval Date :

Decisions:

Reviewer Department Decision Userid Date

9/11/2018 E Course Request

https://act.ucsd.edu/qlink/easyquery?qlink.report.id=ECourseRequest&A=10754 1/3

UNIVERSITY OF CALIFORNIA - SAN DIEGOE COURSE REQUESTDate: 09/11/2018

Request Type :Subject Code :Course Number :

New

MDS

230

Department :Course Title :Transcript Title :Effective Term :Extent of Approval :Summer Only :

Description :

Computer Science & Engineering

Big Data Analytics Using Spark

Big Data Analytics Using Spark

FALL 20

NThis course covers techniques for achieving scalability in data analysis, using tools such as MapReduce, Hadoop and Spark. Covered topics include: Programming Spark using Pyspark; Identifying the computational tradeoffs in a Spark application; Performing data loading and cleaning using Spark and Parquet; Modeling data through statistical and machine learning methods.

Course Justification : Core course of the proposed online Masters of Data Science

Request Justification : Core course of the proposed online Masters of Data Science

FIAT :

Instructors:Instructor Assignment : Department Chair Assigned

Department Title/Rank First Name Initial Last Name

Crosslisted Courses:Subject Code Course Number Department Course Title Status End Term

Instructional Units/Hours:Unit Type : FIXEDFixed Units : 4.00

Variable Units : Min : 0.00 Max : 0.00 Increment : 0.00

Types of Instruction: Fixed Hours Var Hours Min Var Hours Max Grade Report OtherDI 1.00 0.00 0.00 YLE 3.00 0.00 0.00 NOP 8.00 0.00 0.00 N

Total Hours(Fixed) : 12.00 Total(Var)Min/Max :

Course Repeatability:Number of Times Taken forCredit: 1

Total Unit Credits : 4.00Justification :

Grade Options:Undergraduate :Graduate : Standard(letter/or S/U)

9/11/2018 E Course Request

https://act.ucsd.edu/qlink/easyquery?qlink.report.id=ECourseRequest&A=10754 2/3

In Progress Grading:Does course use in-progressgrading : N

IP Justification :

In Progress Sequence Courses:

Status Subject Code Course No. Course Title Seq No. Final Grade End Term

Final Evaluation:Final Exam Method : P3/O

Keys: C=In Class Final Exam; L=Lab Final; P=Final Paper; P2=Final Presentation; P3=Final Project; O=Other; NL=Nofinal(Lab); NG=No Final(Grad Course);

Description : remotely proctored final exam

Justification : Final project, in conjunction with a remotely proctored final exam,represents the optimal method for final evaluation for this course

Course Prerequisites:

Status Department Subject Code Course No. Title End Term Prerequisite Seq

Test Prerequisites:

Test Description Low Score High Score

Department Restrictions:Department Approval Req : NDepartment Approval Req Justification:Other Enrollment Requirements : Restricted to students within the MDS major code.

Academic Level Restrictions:

Freshman: N Sophomore : N Junior : N Senior : NLower Div: N Upper Div : N Graduate : N Pharmacy : N Medical : N

Course Corequisites:

Status Department Subject Code Course No. Course Title End Term

Course Duplicates:

Status Department Subject Code Course No. Course Title End Term

Other Catalog Information:Other Catalog Information(Optional):Recommended Preporation(Optional) :Material Fee : N

Animal Subjects:Uses Animal Subjects: NProtocol Number :Approval Date :

Human Subjects:Uses Human Subjects: NProtocol Number :Approval Date :

9/11/2018 E Course Request

https://act.ucsd.edu/qlink/easyquery?qlink.report.id=ECourseRequest&A=10754 3/3

Decisions:

Reviewer Department Decision Userid Date

9/11/2018 E Course Request

https://act.ucsd.edu/qlink/easyquery?qlink.report.id=ECourseRequest&A=10758 1/3

UNIVERSITY OF CALIFORNIA - SAN DIEGOE COURSE REQUESTDate: 09/11/2018

Request Type :Subject Code :Course Number :

New

MDS

240

Department :Course Title :Transcript Title :Effective Term :Extent of Approval :Summer Only :

Description :

Computer Science & Engineering

Data Mining on the Web

Data Mining on the Web

FALL 20

NThis course covers recommender systems, data mining, and predictive analytics. Students will learn how to build models that help us understand data in order to gain insights and make predictions. All programming assignments are in Python. Covered topics include: Regression; Classification; Unsupervised learning and dimensionality reduction; Recommender systems; Text mining; Social network analysis; Visualization; Crawling; Online advertising.

Course Justification : Core course of the proposed online Masters of Data Science

Request Justification : Core course of the proposed online Masters of Data Science

FIAT :

Instructors:Instructor Assignment : Department Chair Assigned

Department Title/Rank First Name Initial Last Name

Crosslisted Courses:Subject Code Course Number Department Course Title Status End Term

Instructional Units/Hours:Unit Type : FIXEDFixed Units : 4.00

Variable Units : Min : 0.00 Max : 0.00 Increment : 0.00

Types of Instruction: Fixed Hours Var Hours Min Var Hours Max Grade Report OtherDI 1.00 0.00 0.00 YLE 3.00 0.00 0.00 NOP 8.00 0.00 0.00 N

Total Hours(Fixed) : 12.00 Total(Var)Min/Max :

Course Repeatability:Number of Times Taken forCredit: 1

Total Unit Credits : 4.00Justification :

Grade Options:Undergraduate :Graduate : Standard(letter/or S/U)

9/11/2018 E Course Request

https://act.ucsd.edu/qlink/easyquery?qlink.report.id=ECourseRequest&A=10758 2/3

In Progress Grading:Does course use in-progressgrading : N

IP Justification :

In Progress Sequence Courses:

Status Subject Code Course No. Course Title Seq No. Final Grade End Term

Final Evaluation:Final Exam Method : P3/O

Keys: C=In Class Final Exam; L=Lab Final; P=Final Paper; P2=Final Presentation; P3=Final Project; O=Other; NL=Nofinal(Lab); NG=No Final(Grad Course);

Description : remotely proctored final exam

Justification : Final project, in conjunction with a remotely proctored final exam,represents the optimal method for final evaluation for this course

Course Prerequisites:

Status Department Subject Code Course No. Title End Term Prerequisite Seq

Test Prerequisites:

Test Description Low Score High Score

Department Restrictions:Department Approval Req : NDepartment Approval Req Justification:Other Enrollment Requirements : Restricted to students within the MDS major code.

Academic Level Restrictions:

Freshman: N Sophomore : N Junior : N Senior : NLower Div: N Upper Div : N Graduate : N Pharmacy : N Medical : N

Course Corequisites:

Status Department Subject Code Course No. Course Title End Term

Course Duplicates:

Status Department Subject Code Course No. Course Title End Term

Other Catalog Information:Other Catalog Information(Optional):Recommended Preporation(Optional) :Material Fee : N

Animal Subjects:Uses Animal Subjects: NProtocol Number :Approval Date :

Human Subjects:Uses Human Subjects: NProtocol Number :

9/11/2018 E Course Request

https://act.ucsd.edu/qlink/easyquery?qlink.report.id=ECourseRequest&A=10758 3/3

Approval Date :

Decisions:

Reviewer Department Decision Userid Date

9/11/2018 E Course Request

https://act.ucsd.edu/qlink/easyquery?qlink.report.id=ECourseRequest&A=10759 1/3

UNIVERSITY OF CALIFORNIA - SAN DIEGOE COURSE REQUESTDate: 09/11/2018

Request Type :Subject Code :Course Number :

New

MDS

250

Department :Course Title :Transcript Title :Effective Term :Extent of Approval :Summer Only :

Description :

Computer Science & Engineering

Data Management for Analytics

Data Management for Analytics

FALL 20

NThis course covers how to store and manage large amounts of data, with an eye toward Data Science. The course will introduce a variety of data formats, data models, high-level query languages and programming abstractions relevant for Data Science. Topics include: relational database systems; hierarchical graph database systems; unrestricted graph database systems; array databases; parallel programming abstractions including Map/Reduce and its descendants developed for graph data.

Course Justification : Core course of the proposed online Masters of Data Science.

Request Justification : Core course of the proposed online Masters of Data Science.

FIAT :

Instructors:Instructor Assignment : Department Chair Assigned

Department Title/Rank First Name Initial Last Name

Crosslisted Courses:Subject Code Course Number Department Course Title Status End Term

Instructional Units/Hours:Unit Type : FIXEDFixed Units : 4.00

Variable Units : Min : 0.00 Max : 0.00 Increment : 0.00

Types of Instruction: Fixed Hours Var Hours Min Var Hours Max Grade Report OtherDI 1.00 0.00 0.00 YLE 3.00 0.00 0.00 NOP 8.00 0.00 0.00 N

Total Hours(Fixed) : 12.00 Total(Var)Min/Max :

Course Repeatability:Number of Times Taken forCredit: 1

Total Unit Credits : 4.00Justification :

Grade Options:Undergraduate :

9/11/2018 E Course Request

https://act.ucsd.edu/qlink/easyquery?qlink.report.id=ECourseRequest&A=10759 2/3

Graduate : Standard(letter/or S/U)

In Progress Grading:Does course use in-progressgrading : N

IP Justification :

In Progress Sequence Courses:

Status Subject Code Course No. Course Title Seq No. Final Grade End Term

Final Evaluation:Final Exam Method : P3/O

Keys: C=In Class Final Exam; L=Lab Final; P=Final Paper; P2=Final Presentation; P3=Final Project; O=Other; NL=Nofinal(Lab); NG=No Final(Grad Course);

Description : remotely proctored final exam

Justification : Final project, in conjunction with a remotely proctored final exam,represents the optimal method for final evaluation for this course.

Course Prerequisites:

Status Department Subject Code Course No. Title End Term Prerequisite Seq

Test Prerequisites:

Test Description Low Score High Score

Department Restrictions:Department Approval Req : NDepartment Approval Req Justification:Other Enrollment Requirements : Restricted to students within the MDS major code.

Academic Level Restrictions:

Freshman: N Sophomore : N Junior : N Senior : NLower Div: N Upper Div : N Graduate : N Pharmacy : N Medical : N

Course Corequisites:

Status Department Subject Code Course No. Course Title End Term

Course Duplicates:

Status Department Subject Code Course No. Course Title End Term

Other Catalog Information:Other Catalog Information(Optional):Recommended Preporation(Optional) :Material Fee : N

Animal Subjects:Uses Animal Subjects: NProtocol Number :Approval Date :

Human Subjects:Uses Human Subjects: N

9/11/2018 E Course Request

https://act.ucsd.edu/qlink/easyquery?qlink.report.id=ECourseRequest&A=10759 3/3

Protocol Number :Approval Date :

Decisions:

Reviewer Department Decision Userid Date

9/11/2018 E Course Request

https://act.ucsd.edu/qlink/easyquery?qlink.report.id=ECourseRequest&A=10755 1/3

UNIVERSITY OF CALIFORNIA - SAN DIEGOE COURSE REQUESTDate: 09/11/2018

Request Type :Subject Code :Course Number :

New

MDS

260

Department :Course Title :Transcript Title :Effective Term :Extent of Approval :Summer Only :

Description :

Computer Science & Engineering

Advanced Unsupervised Learning

Advanced Unsupervised Learning

WINTER 21

NThis course covers advanced techniques for unsupervised learning. Covered topics include: Dimensionality reduction; k-means; Principal component analysis; Topic modeling; Deep unsupervised models.

Course Justification : Elective course of the proposed online Masters of Data Science.

Request Justification : Elective course of the proposed online Masters of Data Science.

FIAT :

Instructors:Instructor Assignment : Department Chair Assigned

Department Title/Rank First Name Initial Last Name

Crosslisted Courses:Subject Code Course Number Department Course Title Status End Term

Instructional Units/Hours:Unit Type : FIXEDFixed Units : 4.00

Variable Units : Min : 0.00 Max : 0.00 Increment : 0.00

Types of Instruction: Fixed Hours Var Hours Min Var Hours Max Grade Report OtherDI 1.00 0.00 0.00 YLE 3.00 0.00 0.00 NOP 8.00 0.00 0.00 N

Total Hours(Fixed) : 12.00 Total(Var)Min/Max :

Course Repeatability:Number of Times Taken forCredit: 1

Total Unit Credits : 4.00Justification :

Grade Options:Undergraduate :Graduate : Standard(letter/or S/U)

In Progress Grading:

9/11/2018 E Course Request

https://act.ucsd.edu/qlink/easyquery?qlink.report.id=ECourseRequest&A=10755 2/3

Does course use in-progressgrading :

N

IP Justification :

In Progress Sequence Courses:

Status Subject Code Course No. Course Title Seq No. Final Grade End Term

Final Evaluation:Final Exam Method : P3/O

Keys: C=In Class Final Exam; L=Lab Final; P=Final Paper; P2=Final Presentation; P3=Final Project; O=Other; NL=Nofinal(Lab); NG=No Final(Grad Course);

Description : remotely proctored final exam

Justification : Final project, in conjunction with a remotely proctored final exam,represents the optimal method for final evaluation for this course

Course Prerequisites:

Status Department Subject Code Course No. Title End Term Prerequisite Seq

Test Prerequisites:

Test Description Low Score High Score

Department Restrictions:Department Approval Req : NDepartment Approval Req Justification:Other Enrollment Requirements : Restricted to students within the MDS major code.

Academic Level Restrictions:Freshman: N Sophomore : N Junior : N Senior : NLower Div: N Upper Div : N Graduate : N Pharmacy : N Medical : N

Course Corequisites:

Status Department Subject Code Course No. Course Title End Term

Course Duplicates:

Status Department Subject Code Course No. Course Title End Term

Other Catalog Information:Other Catalog Information(Optional):Recommended Preporation(Optional) :Material Fee : N

Animal Subjects:Uses Animal Subjects: NProtocol Number :Approval Date :

Human Subjects:Uses Human Subjects: NProtocol Number :Approval Date :

9/11/2018 E Course Request

https://act.ucsd.edu/qlink/easyquery?qlink.report.id=ECourseRequest&A=10755 3/3

Decisions:

Reviewer Department Decision Userid Date

9/11/2018 E Course Request

https://act.ucsd.edu/qlink/easyquery?qlink.report.id=ECourseRequest&A=10756 1/3

UNIVERSITY OF CALIFORNIA - SAN DIEGOE COURSE REQUESTDate: 09/11/2018

Request Type :Subject Code :Course Number :

New

MDS

261

Department :Course Title :Transcript Title :Effective Term :Extent of Approval :Summer Only :

Description :

Computer Science & Engineering

From Data to Products

From Data to Products

WINTER 21

NThis course teaches students how to build an entire data-processing pipeline that can be used in production. Covered topics include: Mechanisms for putting machine learning techniques into production; Data cleaning; Data visualization.

Course Justification : Elective course of the proposed online Masters of Data Science.

Request Justification : Elective course of the proposed online Masters of Data Science.

FIAT :

Instructors:Instructor Assignment : Department Chair Assigned

Department Title/Rank First Name Initial Last Name

Crosslisted Courses:Subject Code Course Number Department Course Title Status End Term

Instructional Units/Hours:Unit Type : FIXEDFixed Units : 4.00

Variable Units : Min : 0.00 Max : 0.00 Increment : 0.00

Types of Instruction: Fixed Hours Var Hours Min Var Hours Max Grade Report OtherDI 1.00 0.00 0.00 YLE 3.00 0.00 0.00 NOP 8.00 0.00 0.00 N

Total Hours(Fixed) : 12.00 Total(Var)Min/Max :

Course Repeatability:Number of Times Taken forCredit: 1

Total Unit Credits : 4.00Justification :

Grade Options:Undergraduate :Graduate : Standard(letter/or S/U)

In Progress Grading:

9/11/2018 E Course Request

https://act.ucsd.edu/qlink/easyquery?qlink.report.id=ECourseRequest&A=10756 2/3

Does course use in-progressgrading :

N

IP Justification :

In Progress Sequence Courses:

Status Subject Code Course No. Course Title Seq No. Final Grade End Term

Final Evaluation:Final Exam Method : P3/O

Keys: C=In Class Final Exam; L=Lab Final; P=Final Paper; P2=Final Presentation; P3=Final Project; O=Other; NL=Nofinal(Lab); NG=No Final(Grad Course);

Description : remotely proctored final exam

Justification : Final project, in conjunction with a remotely proctored final exam,represents the optimal method for final evaluation for this course

Course Prerequisites:

Status Department Subject Code Course No. Title End Term Prerequisite Seq

Test Prerequisites:

Test Description Low Score High Score

Department Restrictions:Department Approval Req : NDepartment Approval Req Justification:Other Enrollment Requirements : Restricted to students within the MDS major code.

Academic Level Restrictions:

Freshman: N Sophomore : N Junior : N Senior : NLower Div: N Upper Div : N Graduate : N Pharmacy : N Medical : N

Course Corequisites:

Status Department Subject Code Course No. Course Title End Term

Course Duplicates:

Status Department Subject Code Course No. Course Title End Term

Other Catalog Information:Other Catalog Information(Optional):Recommended Preporation(Optional) :Material Fee : N

Animal Subjects:Uses Animal Subjects: NProtocol Number :Approval Date :

Human Subjects:Uses Human Subjects: NProtocol Number :Approval Date :

9/11/2018 E Course Request

https://act.ucsd.edu/qlink/easyquery?qlink.report.id=ECourseRequest&A=10756 3/3

Decisions:

Reviewer Department Decision Userid Date

9/11/2018 E Course Request

https://act.ucsd.edu/qlink/easyquery?qlink.report.id=ECourseRequest&A=10757 1/3

UNIVERSITY OF CALIFORNIA - SAN DIEGOE COURSE REQUESTDate: 09/11/2018

Request Type :Subject Code :Course Number :

New

MDS

262

Department :Course Title :Transcript Title :Effective Term :Extent of Approval :Summer Only :

Description :

Computer Science & Engineering

Data Visualization

Data Visualization

WINTER 21

NThe course covers techniques for creating effective visualizations to explore trends, identify relationships, confirm hypotheses, communicate findings and gain insight about data. This course will focus on teaching students the principles and techniques for creating visual representation from raw data. The course exercises will be based on publicly available datasets and utilize freely available tools like D3.js.

Course Justification : Elective course of the proposed online Masters of Data Science

Request Justification : Elective course of the proposed online Masters of Data Science

FIAT :

Instructors:Instructor Assignment : Department Chair Assigned

Department Title/Rank First Name Initial Last Name

Crosslisted Courses:Subject Code Course Number Department Course Title Status End Term

Instructional Units/Hours:Unit Type : FIXEDFixed Units : 4.00

Variable Units : Min : 0.00 Max : 0.00 Increment : 0.00

Types of Instruction: Fixed Hours Var Hours Min Var Hours Max Grade Report OtherDI 1.00 0.00 0.00 YLE 3.00 0.00 0.00 NOP 8.00 0.00 0.00 N

Total Hours(Fixed) : 12.00 Total(Var)Min/Max :

Course Repeatability:Number of Times Taken forCredit: 1

Total Unit Credits : 4.00Justification :

Grade Options:Undergraduate :Graduate : Standard(letter/or S/U)

9/11/2018 E Course Request

https://act.ucsd.edu/qlink/easyquery?qlink.report.id=ECourseRequest&A=10757 2/3

In Progress Grading:Does course use in-progressgrading : N

IP Justification :

In Progress Sequence Courses:

Status Subject Code Course No. Course Title Seq No. Final Grade End Term

Final Evaluation:Final Exam Method : P3/O

Keys: C=In Class Final Exam; L=Lab Final; P=Final Paper; P2=Final Presentation; P3=Final Project; O=Other; NL=Nofinal(Lab); NG=No Final(Grad Course);

Description : remotely proctored final exam

Justification : Final project, in conjunction with a remotely proctored final exam,represents the optimal method for final evaluation for this course

Course Prerequisites:

Status Department Subject Code Course No. Title End Term Prerequisite Seq

Test Prerequisites:

Test Description Low Score High Score

Department Restrictions:Department Approval Req : NDepartment Approval Req Justification:Other Enrollment Requirements : Restricted to students within the MDS major code.

Academic Level Restrictions:

Freshman: N Sophomore : N Junior : N Senior : NLower Div: N Upper Div : N Graduate : N Pharmacy : N Medical : N

Course Corequisites:

Status Department Subject Code Course No. Course Title End Term

Course Duplicates:

Status Department Subject Code Course No. Course Title End Term

Other Catalog Information:Other Catalog Information(Optional):Recommended Preporation(Optional) :Material Fee : N

Animal Subjects:Uses Animal Subjects: NProtocol Number :Approval Date :

Human Subjects:Uses Human Subjects: NProtocol Number :

9/11/2018 E Course Request

https://act.ucsd.edu/qlink/easyquery?qlink.report.id=ECourseRequest&A=10757 3/3

Approval Date :

Decisions:

Reviewer Department Decision Userid Date

9/11/2018 E Course Request

https://act.ucsd.edu/qlink/easyquery?qlink.report.id=ECourseRequest&A=10760 1/3

UNIVERSITY OF CALIFORNIA - SAN DIEGOE COURSE REQUESTDate: 09/11/2018

Request Type :Subject Code :Course Number :

New

MDS

263

Department :Course Title :Transcript Title :Effective Term :Extent of Approval :Summer Only :

Description :

Computer Science & Engineering

Data Preprocessing

Data Preprocessing

SPRING 21

NThis courses covers techniques for taking raw data (for example collected directly from web pages) and converting it into a clean data set that machine learning techniques can work on. The goal of this course is to understand the nature of information heterogeneity, the techniques of relating information from different sources, and the machinery required for achieving the integration.

Course Justification : Elective course of the proposed online Masters of Data Science.

Request Justification : Elective course of the proposed online Masters of Data Science.

FIAT :

Instructors:Instructor Assignment : Department Chair Assigned

Department Title/Rank First Name Initial Last Name

Crosslisted Courses:Subject Code Course Number Department Course Title Status End Term

Instructional Units/Hours:Unit Type : FIXEDFixed Units : 4.00

Variable Units : Min : 0.00 Max : 0.00 Increment : 0.00

Types of Instruction: Fixed Hours Var Hours Min Var Hours Max Grade Report OtherDI 1.00 0.00 0.00 YLE 3.00 0.00 0.00 NOP 8.00 0.00 0.00 N

Total Hours(Fixed) : 12.00 Total(Var)Min/Max :

Course Repeatability:Number of Times Taken forCredit: 1

Total Unit Credits : 4.00Justification :

Grade Options:Undergraduate :Graduate : Standard(letter/or S/U)

9/11/2018 E Course Request

https://act.ucsd.edu/qlink/easyquery?qlink.report.id=ECourseRequest&A=10760 2/3

In Progress Grading:Does course use in-progressgrading : N

IP Justification :

In Progress Sequence Courses:

Status Subject Code Course No. Course Title Seq No. Final Grade End Term

Final Evaluation:Final Exam Method : P3/O

Keys: C=In Class Final Exam; L=Lab Final; P=Final Paper; P2=Final Presentation; P3=Final Project; O=Other; NL=Nofinal(Lab); NG=No Final(Grad Course);

Description : remotely proctored final exam

Justification : Final project, in conjunction with a remotely proctored final exam,represents the optimal method for final evaluation for this course.

Course Prerequisites:

Status Department Subject Code Course No. Title End Term Prerequisite Seq

Test Prerequisites:

Test Description Low Score High Score

Department Restrictions:Department Approval Req : NDepartment Approval Req Justification:Other Enrollment Requirements : Restricted to students within the MDS major code.

Academic Level Restrictions:

Freshman: N Sophomore : N Junior : N Senior : NLower Div: N Upper Div : N Graduate : N Pharmacy : N Medical : N

Course Corequisites:

Status Department Subject Code Course No. Course Title End Term

Course Duplicates:

Status Department Subject Code Course No. Course Title End Term

Other Catalog Information:Other Catalog Information(Optional):Recommended Preporation(Optional) :Material Fee : N

Animal Subjects:Uses Animal Subjects: NProtocol Number :Approval Date :

Human Subjects:Uses Human Subjects: NProtocol Number :Approval Date :

9/11/2018 E Course Request

https://act.ucsd.edu/qlink/easyquery?qlink.report.id=ECourseRequest&A=10760 3/3

Decisions:

Reviewer Department Decision Userid Date

9/11/2018 E Course Request

https://act.ucsd.edu/qlink/easyquery?qlink.report.id=ECourseRequest&A=10761 1/3

UNIVERSITY OF CALIFORNIA - SAN DIEGOE COURSE REQUESTDate: 09/11/2018

Request Type :Subject Code :Course Number :

New

MDS

264

Department :Course Title :Transcript Title :Effective Term :Extent of Approval :Summer Only :

Description :

Computer Science & Engineering

Interaction Design

Interaction Design

SPRING 21

NUsers interact with data through user interfaces. This course introduces fundamental methods and principles for designing, implementing, and evaluating user interfaces. Students will learn how to generate design ideas, techniques for prototyping them, and how to use prototypes to get feedback from stakeholders. Topics include: user-centered design, rapid prototyping, experimentation, direct manipulation, cognitive principles, visual design, social software, software tools.

Course Justification : Elective course of the proposed online Masters of Data Science.

Request Justification : Elective course of the proposed online Masters of Data Science.

FIAT :

Instructors:Instructor Assignment : Department Chair Assigned

Department Title/Rank First Name Initial Last Name

Crosslisted Courses:Subject Code Course Number Department Course Title Status End Term

Instructional Units/Hours:Unit Type : FIXEDFixed Units : 4.00

Variable Units : Min : 0.00 Max : 0.00 Increment : 0.00

Types of Instruction: Fixed Hours Var Hours Min Var Hours Max Grade Report OtherDI 1.00 0.00 0.00 YLE 3.00 0.00 0.00 NOP 8.00 0.00 0.00 N

Total Hours(Fixed) : 12.00 Total(Var)Min/Max :

Course Repeatability:Number of Times Taken forCredit: 1

Total Unit Credits : 4.00Justification :

Grade Options:Undergraduate :Graduate : Standard(letter/or S/U)

9/11/2018 E Course Request

https://act.ucsd.edu/qlink/easyquery?qlink.report.id=ECourseRequest&A=10761 2/3

In Progress Grading:Does course use in-progressgrading : N

IP Justification :

In Progress Sequence Courses:

Status Subject Code Course No. Course Title Seq No. Final Grade End Term

Final Evaluation:Final Exam Method : P3/O

Keys: C=In Class Final Exam; L=Lab Final; P=Final Paper; P2=Final Presentation; P3=Final Project; O=Other; NL=Nofinal(Lab); NG=No Final(Grad Course);

Description : remotely proctored final exam

Justification : Final project, in conjunction with a remotely proctored final exam,represents the optimal method for final evaluation for this course.

Course Prerequisites:

Status Department Subject Code Course No. Title End Term Prerequisite Seq

Test Prerequisites:

Test Description Low Score High Score

Department Restrictions:Department Approval Req : NDepartment Approval Req Justification:Other Enrollment Requirements : Restricted to students within the MDS major code.

Academic Level Restrictions:

Freshman: N Sophomore : N Junior : N Senior : NLower Div: N Upper Div : N Graduate : N Pharmacy : N Medical : N

Course Corequisites:

Status Department Subject Code Course No. Course Title End Term

Course Duplicates:

Status Department Subject Code Course No. Course Title End Term

Other Catalog Information:Other Catalog Information(Optional):Recommended Preporation(Optional) :Material Fee : N

Animal Subjects:Uses Animal Subjects: NProtocol Number :Approval Date :

Human Subjects:Uses Human Subjects: NProtocol Number :

9/11/2018 E Course Request

https://act.ucsd.edu/qlink/easyquery?qlink.report.id=ECourseRequest&A=10761 3/3

Approval Date :

Decisions:

Reviewer Department Decision Userid Date

9/11/2018 E Course Request

https://act.ucsd.edu/qlink/easyquery?qlink.report.id=ECourseRequest&A=10762 1/3

UNIVERSITY OF CALIFORNIA - SAN DIEGOE COURSE REQUESTDate: 09/11/2018

Request Type :Subject Code :Course Number :

New

MDS

298

Department :Course Title :Transcript Title :Effective Term :Extent of Approval :Summer Only :

Description :

Computer Science & Engineering

Capstone Project in Data Science

Capstone Project in DataSci

SPRING 21

NThis course consists of a quarter-long project, taken from several possible domains, including: Music, Oceanography, and Computer Vision. The project will require students to apply the material they learned throughout the program to practice.

Course Justification : Capstone course of the proposed online Masters of Data Science.

Request Justification : Capstone course of the proposed online Masters of Data Science.

FIAT :

Instructors:Instructor Assignment : Department Chair Assigned

Department Title/Rank First Name Initial Last Name

Crosslisted Courses:Subject Code Course Number Department Course Title Status End Term

Instructional Units/Hours:Unit Type : FIXEDFixed Units : 4.00

Variable Units : Min : 0.00 Max : 0.00 Increment : 0.00

Types of Instruction: Fixed Hours Var Hours Min Var Hours Max Grade Report OtherDI 1.00 0.00 0.00 YLE 3.00 0.00 0.00 NOP 8.00 0.00 0.00 N

Total Hours(Fixed) : 12.00 Total(Var)Min/Max :

Course Repeatability:Number of Times Taken forCredit: 1

Total Unit Credits : 4.00Justification :

Grade Options:Undergraduate :Graduate : Standard(letter/or S/U)

9/11/2018 E Course Request

https://act.ucsd.edu/qlink/easyquery?qlink.report.id=ECourseRequest&A=10762 2/3

In Progress Grading:Does course use in-progressgrading : N

IP Justification :

In Progress Sequence Courses:

Status Subject Code Course No. Course Title Seq No. Final Grade End Term

Final Evaluation:Final Exam Method : P3

Keys: C=In Class Final Exam; L=Lab Final; P=Final Paper; P2=Final Presentation; P3=Final Project; O=Other; NL=Nofinal(Lab); NG=No Final(Grad Course);

Description :

Justification : The project will require students to apply the material they learnedthroughout the program to practice.

Course Prerequisites:

Status Department Subject Code Course No. Title End Term Prerequisite Seq

Test Prerequisites:

Test Description Low Score High Score

Department Restrictions:Department Approval Req : NDepartment Approval Req Justification:Other Enrollment Requirements : Restricted to students within the MDS major code.

Academic Level Restrictions:

Freshman: N Sophomore : N Junior : N Senior : NLower Div: N Upper Div : N Graduate : N Pharmacy : N Medical : N

Course Corequisites:

Status Department Subject Code Course No. Course Title End Term

Course Duplicates:

Status Department Subject Code Course No. Course Title End Term

Other Catalog Information:Other Catalog Information(Optional):Recommended Preporation(Optional) :Material Fee : N

Animal Subjects:Uses Animal Subjects: NProtocol Number :Approval Date :

Human Subjects:Uses Human Subjects: NProtocol Number :Approval Date :

9/11/2018 E Course Request

https://act.ucsd.edu/qlink/easyquery?qlink.report.id=ECourseRequest&A=10762 3/3

Decisions:

Reviewer Department Decision Userid Date

Response to Graduate Council We thank the Graduate Council for the new guidelines for online graduate programs. In this document we:

● Outline the main changes since we last submitted ● We provide pointers to where in the proposal we address each one of questions from the

guidelines ● We again answer prior questions that the Graduate Council had, in some cases with

updated information, in some cases with the same answer as before. At the end of this document, we attach a new version of the body of the proposal with track changes turned on so that changes are easy to see. Finally, in a separate document, we provide a complete clean new proposal packet, with the new proposal body.

Main Changes ● As required by the new UCSD policy on cooperation with EdX, we have removed all ties

to edX, except for the use of the MicroMasters as a pathway. The program will not be offered with edX anymore.

● After considering other online programs, especially ones started in the past few years, we saw that many programs are in the 20-30K, and so we adjusted the price of the degree to a total of $22,000. So we have a new budget.

● We consolidated and expanded the discussion of Diversity, Equity and Inclusion in a new section, 1.5

● We consolidated and expanded the evaluation of the program in Section 1.8 to address many of the guidelines provided by Graduate Council

● We expanded 2.3.8 to further discuss quality of education in an online setting

Answers to questions from guidelines I Rationale 1. Within the stated aims and objectives of the program, clearly state: (1) the educational rationale for offering the proposed program in an online format; (2) the learning outcomes for the online degree program; (3) the target audience for the online degree program, including how the online format will benefit the target audience; (4) the core design characteristics of the program (e.g., the relationship between program size and program cost). Please see Section 1.1 of the proposal.

2. If the proposed online program augments or replaces an existing brick and mortar program, explain how the change to an online format will benefit the educational mission of UC San Diego. Please see Sections 1.1, 1.2, and 1.4 of the proposal. 3. Please explain how future revenues from the online Master’s Degree program will be used to benefit the educational mission of UC San Diego. Please see Sections 1.5, 6.9, and 7 of the proposal. II. Faculty Incentives and Compensation 1. Provide specific details about faculty incentive packages for program and course development. Please see Section 6.1 of the proposal. a. Explain how compensation for course creators aligns with the “Guidelines for Revenue Sharing Payments Related to Executed Online Course Development Agreements” (prepared by the Office of the Executive Vice Chancellor). Link: https://digitallearning.ucsd.edu/_files/Online-Course-Development-Guidelines---A ugust-2019.pdf Please see Section 6.1 of the proposal. b. Describe how faculty incentives will favor quality course design. Please see Sections 2.3.7 and 6.1 of the proposal. 2. Provide specific details about how faculty will be compensated for program oversight and course instruction. Please see Section 6.1 of the proposal. a. Provide the plans for who will be teaching the courses for the program. Please see Sections 4, 5.2, 5.3, 5.4 and 5.5. b. If an SSGPDP, specify whether courses will be assigned as part of the instructors’ assigned regular teaching load or overload. Please refer to the UCSD Guidelines for Teaching in Self-Supporting Programs: https://aps.ucsd.edu/_files/compensation/overloadguidelines.pdf and illustrate

how the plan for teaching assignments, and revenue use, will be in compliance with the guidelines. Please see Section 6.1 3. Explain how faculty compensation for teaching in the degree program reflects and supports the needs of the academic unit and does not diminish the academic unit’s responsibilities to its full complement of state-supported and on-campus programs. Please see Section 6.1, specifically under “Faculty participation in state-supported programs” 4. Explain how revenue will be returned to the department (and/or division), including to support new or existing faculty FTEs, in accordance with the UC San Diego guidelines: http://academicaffairs.ucsd.edu/sso/SAPD/Financial/ . (Note: This link to the guidelines is subject to change; this Graduate Council document will be updated with the correct link on the Senate website if the link changes) Please see Sections 6.1 and 6.9. III. Quality Course Design 1. Identify the program’s requirements for campus-level resources to support instructional and learning needs and provide evidence that the identified support will be provided. The explicit budget costs of these non-faculty instructional personnel should be shown. Please see Sections 6.1 and 6.6. 2. Explain what learning management systems and digital technologies will be utilized and how students and instructors will engage with them. Describe what institutional resources are required to support these plans and provide documentation of the institution’s commitment to provide support. Identify what divisions and personnel will provide this support. Please see Sections 5.1, 6.6, and the letters of support. 3. Provide a rationale for the instructional design of the proposed program. All proposals should describe the level of engagement and consultation between the program (including individual faculty) and Digital Learning in the Teaching + Learning Commons. Please see section 2.3.7. 4. Explain how scalable forms of feedback and guidance (e.g., TAs, peer review, automation, adaptive learning, etc.) will provide quality learning and assessment experiences for the target student population.

Please see Section 5.1. 5. Identify all student support services required for the online degree program, and explain how these will be provided and resourced. Please see Section 8. IV. Faculty and Teaching Assistant Involvement 1. Provide the names and roles of the faculty involved with each of the following: program design, course development, and course instruction. Describe their qualifications for developing and/or teaching in an online degree program. Please see Sections 2.3.7, 4, 5.2, 5.3, 5.4, and 5.5. 2. Describe what approaches will be taken to support faculty-student interaction, student-teaching assistant (TA) interaction, student-student interaction, and instructor-TA interactions. Please see Section 5.1. 3. Describe the proposing academic unit’s expectations of faculty and other teaching staff time commitments evaluated in the context of the academic unit’s needs and resources (especially in consideration of existing on-campus programs). Please see Section 6.1. 4. Describe the proposing academic unit’s processes for recruiting, training, and supporting TAs hired to support online courses. Please see Section 6.5. V. Diversity 1. State the program’s diversity goals in light of the program rationale (see 1. Rationale). Describe how these goals were determined and the resources used or offices consulted to develop the program’s diversity goals. Also provide a plan for assessment of the success of the stated diversity goals. Specifically address what diversity means in an online context and how the proposed online format relates to the proposing academic unit’s overall diversity goals, including its brick and mortar programs. Please see Section 1 and 7.2 of the proposal.

2. Describe specific features of the program’s design, activities, and staffing structure aimed at both recruiting diverse learners and monitoring their success. Please see Section 1.5, 2.1, 2.3.7, 6.9, and 7. 3. Describe how under-represented students will be supported, and whether fellowships aimed at promoting diversity will be deployed. Please see Section 1.5, 6.9 and 7. VI. Graduate Admissions 1. State how graduate admissions will be conducted and staffed in light of possible new student bodies or quantities of applications. Please see Section 2.1 of the proposal. 2. Explain what metrics the program will employ to evaluate student applications and strategies for reducing bias in application evaluation. Please see Section 2.1 of the proposal. VII. Institutional Support: Mitigating Risks to Learners and UC San Diego 1. Describe how intellectual and financial risks to online learners will be mitigated. Please see Section 2.1 of the proposal. 2. Clearly list the institutional resources required to support the online program and provide documentation confirming commitments for institutional support. Specify how much support is required and the responsible academic or administrative unit(s) for providing the support. Please see Section 6 of the proposal and the letters of support. 3. Describe the resources required for supporting the non-academic needs of online learners, including support for withdrawing from classes, academic exceptions, and supporting online learners with disabilities. Include documentation confirming an institutional commitment to support these services. Please see Section 7.1 for the fee and refund policy and Section 7.2 for the needs of online learners. 4. Explain whether and how the academic unit will differentiate the online program from on-campus programs, e.g. adopting “online” in the degree name and describing

programs as “online” on University transcripts. Please see Section 1.1 VIII. Program Assessment 1. Outline the evaluation plan for the program. The recommended evaluation plan for new online degree programs is an annual review during the first five years following establishment. These reviews should include: (1) completion rates for individual classes; (2) completion rates for the degree program; (3) average time to degree; (4) demographic information; (5) each statistic in 1-4 broken down by demographic category; (6) course, TA, and professor response surveys, including student comments and peer instruction reviews; (7) an evaluation from Digital Learning in the Teaching + Learning Commons; and (8) financial data (if an SSGPDP). Please see Section 1.8 and Section 2.3.7. 2. Explain how online classes, instructors, and TAs will be evaluated for both content and format. Please see Section 2.3.7. 3. Describe criteria, as concretely as possible, under which an academic unit would discontinue the online program – e.g. enrollment criteria, time to degree, class completion rate, degree completion rate, diversity, faculty involvement, etc. Please see Section 1.7 and 2.3.7 of the proposal.

Answers to prior Graduate Council questions, revisited 1. Degree Name: A persistent concern among many Graduate Council members is that the degree name “Master of Data Science” does not adequately differentiate the online degree from on-campus degree offerings at UC San Diego. Recommendation: The Graduate Council learned that it is possible for the degree title to include the word “online” - e.g., Master of Data Science (online) without requiring special approvals beyond the standard system wide approvals required to establish a new Master of “X” title for a degree program. We therefore recommend adopting this name for the MDS. The Council confirmed with the Office of the Registrar that additional systemwide approvals are required if the presentation of the degree changes the diploma format, e.g., Online Master of Data Science. We reiterate our previous response:

We have changed the name of the degree to Master of Data Science (online) and changed the proposal to reflect this change. 2. Completion & Time to Degree: The Graduate Council perceives a risk that completion rates and time to degree for an online MDS program may exceed norms of equivalent on-campus professional master’s programs. There are reasons to expect that the MDS may outperform existing MOOC offerings (like the MicroMasters), but there are also reasons to think that a much longer sequence of classes will be harder to complete. Recommendation: The Graduate Council recommends that minimum standards of success are articulated by the proposers relating to completion rates for both individual classes and the entire MDS online degree, below which proposers believe that the reputational costs will be too great to warrant continuation of the program. See revised Sections 1.7 for a detailed response. And also see 2.3.7 for additional information. 3. Financial Impact on Students: In relation to the risk that online offerings exhibit a lower completion rate than on-campus degrees, the Graduate Council perceives a risk that students will overestimate their ability to complete classes, resulting in negative financial impacts. Recommendation: As noted in the Council’s February 11, 2019 memo, the Graduate Council learned that it is possible to alter the standard UCSD refund schedule, to provide a model that reduces risk to an online student population. We recommend adopting such a schedule and articulating it in the proposal. We reiterate our previous response: We wholeheartedly agree with Graduate Council’s recommendation to minimize financial impact on students who may not complete a course. We have adopted the schedule shown below and have updated the proposal accordingly. We follow the same refund schedule as the Georgia Tech Online Masters in Computer Science (a top online program in computing from a top 10 school). Along with our schedule below, we also display the refund that UCSD in-person degrees would give at an equivalent point in time. Our proposed refund schedule is more generous than the standard in-person degree schedule.

% Refund if drop during ... MDS UCSD in Person

Week 1 100% 90%-100%

Week 2 100% 50%

Week 3 70% 25%-50%

Week 4 60% 25%

Week 5 50% 0%-25%

Week 6 40% 0%

Week 7 0% 0%

Week 8 0% 0%

Week 9 0% 0%

Week 10 0% 0%

4. Financial Plan: The Graduate Council has not accessed the proposed contract with edX, and is concerned about any long term commitments such a contract might create. There are also broader questions about how revenues would be used at the department level, and by the campus. Recommendation: We ask that the edX contract be submitted to the Graduate Council, and that any agreed-upon contract allow unilateral exit by UCSD. After admission into the MDS program, students will not be interacting with the edX platform. There is no longer a contract with edX for the MDS degree. The uses of the money are described in Sections 6.1, 6.9, and 7.2. 5. Program Review schedule: Because the MDS program would be the first online degree program if approved, the program will likely benefit from frequent reviews and modifications. Recommendation: The Graduate Council recommends that, if approved, the MDS program commit to submitting an annual report for Graduate Council review for the first five years following establishment. The reports should include: (1) completion rate of individual classes; (2) completion rate of the degree program; (3) average time to degree; (4) demographic information; (5) each statistic in 1-3 broken down by demographic category; (6) course and professor evaluations, similar to those collected for other UCSD classes. The evaluations should include student comments, as is done at the undergraduate level on the CAPE forms. Because this is a self-supporting program, we also request financial data to address whether the MDS program meets financial targets and becomes self-supporting on schedule. The report should also address how funds are deployed on campus and to edX. We reiterate our previous response: We agree with this recommendation. We have updated the proposal to include our commitment to produce such reports. 6. Diversity: Members of Graduate Council had a host of questions regarding the effects of online programs on diversity in higher education, including both the potential of creating online silos, or of serving mainly foreign student populations.

Recommendation: At its March 11 th meeting, the Graduate Council noted that redirecting revenues created from an online MDS toward student scholarships that promote diversity on campus might offset potential concerns regarding the diversity of online learners. Also, we recommend that, when issuing progress reports, the MDS includes an analysis of online demographics, including comparisons with data from the on-campus MAS degree program in Data Science and Engineering. We reiterate our previous response: To offset concerns about diversity, the academic units offering the program will commit to spending at least 25% of the academic unit net revenue for the first three years toward Diversity, Equity and Inclusion efforts for in-person programs (net revenue is the total revenue minus all expenses needed to run the program successfully). A primary example of this would be to create DEI fellowships for in-person programs, say fully funded 5-year PhD fellowships awarded for students based on their contributions to diversity, equity, and inclusion. After three years of operation, a re-assessment will be made regarding the proportion of money that should be used toward DEI efforts, based on data collected during the first three years. We will also follow Graduate Council’s above recommendation about what to include in the progress reports. We have updated the proposal accordingly. 7. Teaching model: The Graduate Council perceives various risks associated with the creation of self-supporting degree programs that feature substantial off-load teaching. Recommendation: The Council recommends that some minimum percentage of classes be taught on-load. We recommend that the program implement a minimum on-load percentage for teaching. We reiterate our previous response: On-load versus off-load teaching is an important issue that requires constant monitoring and adjustments as needed. As a matter of principle, we are committed to ensuring excellence in instruction, all taught with senate faculty. Indeed, our goal will be to have all courses taught by senate faculty, from academic units across campus that participate in Data Science. However, to achieve this goal, the program needs some flexibility when it comes to onload vs offload. Indeed, the exact ratio of onload vs offload is difficult to predict because of teaching workload variations across participating units in the program. In some cases, offload teaching might be the only way to have senate faculty meaningfully participate in the program without affecting state-supported programs. In other cases, onload teaching would work perfectly. Ultimately, this is a resource management issue to balance all teaching obligations across different units and different programs, while maximizing participation by senate faculty. We ask Graduate Council to grant some flexibility in reaching a balance between onload and offload teaching that ensures excellent teaching from senate faculty in all of our programs.

Present plans expect 40% of the teaching to be done onload at the beginning. We will aim for the percent of onload teaching to not go below 40%, subject to the constraint of not affecting state-funded programs (which is a UCOP/CCGA requirement). As campus continues to hire more data science faculty, we will eventually be able to do more MDS teaching onload without compromising state-funded programs. Finally, as already stated, department revenue will partly be used to fund additional senate FTEs, which will raise overall teaching capacity (and, thereby, onload percentage) over time. We have updated the proposal accordingly. 8. Academic integrity: The Graduate Council requests a response to the following question from the Council’s December 13, 2018 memo: “With a pre-designed capstone project, how will the Program ensure the integrity of the project from one year to the next? Will the topics for the pre-designed projects change each year? The Council would like to know what safeguards the Program will implement to promote academic integrity and prevent students from accessing capstone materials from previous years.” We reiterate our previous response: Our proposal has an entire section on academic integrity, Section 2.3.6. We expand here on the items that answer the questions above. We have updated the proposal to reflect these answers. First, the program will change the projects over time. Each year, small changes will be made so that solutions from previous years cannot be copied directly. Every few years, a major project overhaul will take place. Second, we will use plagiarism detection tools. We will make use of a particular tool called Moss, which has become a standard for plagiarism detection in computer science. The Computer Science and Engineering department has a lot of experience using Moss for large in-person classes (classes with over 200 students, both at the undergraduate level and at the MS level). Moss takes a set of assignment submissions and compares each submission which all other submissions. It provides a report of similarity for each pair of submissions. Moss is based on a sophisticated algorithm that is not fooled by simple changes made to an assignment. Put differently, if a student copies someone else’s assignment and makes only minor changes, Moss will detect this. To take into account submissions from prior years, we will simply add the submissions from the capstone project of prior years into Moss. We have done precisely this for large in-person classes, and it works well. As mentioned earlier too, the capstones will change over time, and so prior solutions should not work year by year. As mentioned in the proposal, we will also follow best practices as recommended by the UCSD Academic Integrity Office. These practices include: discussing academic integrity in the syllabus and in the first lecture; providing students with a link that explains the appropriate academic integrity policies and processes; clearly stating the grading policy for violations of academic

integrity; carefully documenting all allegations of academic integrity violations; having students sign an Academic Honesty Statement each time they submit an assignment, project, or exam. Finally, as mentioned in the proposal, we will also have a senate faculty who will be charged to serve as Academic Integrity Coordinator for the program. This person will ensure that all of the above measures are followed.

UNIVERSITY OF CALIFORNIA, SAN DIEGO UCSD BERKELEY • DAVIS • IRVINE • LOS ANGELES • MERCED • RIVERSIDE • SAN DIEGO • SAN FRANCISCO SANTA BARBARA • SANTA CRUZ

DEPARTMENT OF COMPUTER SCIENCE & ENGINEERING HALICIOĞLU DATA SCIENCE INSTITUTE

August 4, 2020 To: Russell Lynn Chair, Graduate Council By mail: Lori Hullings Dear Professor Lynn Thank you for your response of June 25 2020 re Online Master of Data Science. We are heartened by positive reception by the committee. This letter will clarify our response to two questions posed in your letter. To what degree does the content of courses designed for the online MDS program overlap with existing courses offered in participating departments?

In short, the courses provided in the online MDS program are unique in both their format and their focus on practical and interdisciplinary applications. While there are courses in CSE, ECE and Cognitive Science that teach machine learning and artificial intelligence, the MDS courses will present the material from the ground up, in a re-imagined way to focus on the Data Science discipline. If there is some small amount of overlap, this overlap will be valued because it will allow us to leverage our current experience with in-person content to develop online content. At the same time, we note that online content that covers the same topics as an in-person course may differ substantially from the in-person course (e.g., video content, frequent online feedback, etc.) as articulated in the proposal.

Below we provide a more detailed description of how MDS courses are different from prior offerings.

There are four micro masters courses currently offered by the CSE department taught by faculty and researchers across ECE, CSE, and SDSC. These courses are online versions of classes already taught in the Data Science Engineering Master of Advance Science.

1. MDS 200R: Python for Data Science. This course introduces students to several important tools that are needed in Data Science, including Python, Jupyter Notebooks, Pandas, NumPy, Matplotlib, Scikit-learn, and the NLTK Data.

2. MDS 210R: Probability and Statistics in Data Science using Python. This course covers the foundations of probability and statistics needed for Data Science.

3. MDS 220R: Machine Learning Fundamentals. This course covers a variety of basic supervised and unsupervised learning algorithms and the theory behind those algorithms.

4. MDS 230R: Big Data Analytics Using Spark. This course covers practical techniques for doing data analysis on large amounts of data, using tools such as MapReduce, Hadoop, and Spark.

The program will add eight new online courses described below.

1. MDS 240R: Data Mining on the Web. This course covers the application of several

Machine Learning and Data Mining techniques to a variety of applications, including text mining, playlist prediction, suggestion for smart reply, learning visual clothing style, and online advertising.

2. MDS 250R: Data Management for Analytics. This course covers how to store and manage large amounts of data, with an eye toward applications in Data Science.

3. MDS 260R: Advanced Unsupervised Learning. This course covers advanced techniques for unsupervised machine learning, including dimensionality reduction, k-means, principal component analysis, topic modeling and deep unsupervised models.

4. MDS 261R: From Data to Products. This course teaches students how to build an entire data-processing pipeline that can be used in production.

5. MDS 262R: Data Visualization. This course covers techniques for creating effective visualizations to explore trends, identify relationships, test hypotheses, communicate findings and gain insight about data.

6. MDS 263R: Data Preprocessing. This course covers techniques for taking raw data (for example collected directly from web pages) and converting it into a clean data set that machine learning techniques can work on.

7. MDS 264R: Interaction Design. This course introduces fundamental methods and principles for designing, implementing, and evaluating user interfaces for exploring data.

8. MDS 298R: Capstone Project in Data Science. This course consists of a quarter-long project. Students will pick one project out of several available options, each project from a different domain. At launch, we expect to have projects in: Music, Oceanography, and Computer Vision. Over time, we expect to add additional capstone projects from various disciplines, for example, Engineering, Health & Life Sciences, Social Sciences, Physical Sciences, and Arts & Humanities.

There are no existing courses with the exact descriptions as above. While there are some courses in CSE, ECE, and Cognitive Science that cover the general areas of machine learning and artificial intelligence, there are three broad distinctions compared to prior offerings:

(A) The modality of instruction for the Online MDS courses will be entirely different from

other courses, which will lead to a substantially different course. The modality will not only be online (eg: breakup of content in short video sessions, frequent online feedback, interactive sessions, etc), but the course content will also intersperse material with applications in a way that re-imagines traditional material in the context of the new discipline of Data Sciences.

(B) The MDS classes will be geared towards professional students than the courses offered in existing MS programs, which will translate to more applied material.

(C) Since the MDS program puts a significant focus on interdisciplinary material, bringing academic units from across campus together, the classes will have a more interdisciplinary focus.

How will HDSI coordinate with participating departments to prevent duplicating efforts? HDSI faculty council consists of 46 faculty members, 12 of which have 50% or 25% joint appointment with another department on campus covering all divisions including health sciences and marine sciences, in addition to its 11 full-time faculty members, and 19 zero-percent appointments. Together, the HDSI faculty council consists of faculty members who have formal appointments with any and all departments that have any relationship to data science as a subject or application area. During the regular quarter, the faculty council meets weekly for faculty meetings and/or informal talks. This level of engagement and visibility ensures HDSI’s ability to inform and coordinate with participating departments in the online MDS program about any course proposals bidirectionally. More formally, all partially appointed faculty members at HDSI are required to document their annual teaching assignments/commitments in writing that must be reviewed and approved by HDSI and partner departments. Typically courses of mutual interest are jointly listed or cross-listed to ensure proper workload credit for the jointly appointed faculty. This provides a primary means to ensure coordination and prevention of course duplication. The MDS director (who is a senate faculty) will work with Kyle Hofer-Mora, assistant director of HDSI, to coordinate the MDS program, and will meet regularly with participating department committees to provide an update on the program status and plans on a quarterly basis.

If you have any questions or need more information, please do not hesitate to contact us. Sincerely,

Sorin Lerner Professor and Chair Computer Science and Engineering, UC San Diego.

Rajesh K. Gupta Distinguished Professor and Director Halıcıoğlu Data Science Institute, UC San Diego.

Cc: Kyle Hofer, HDSI; JL Morgan: HDSI; Danielle Elias: CSE