the aearu stem summer camp program 2018...5. campbell video1 7.1-7.8 【assignments for this session...

The AEARU STEM Summer Camp Program 2018

The AEARU STEM Summer Camp 2018 will provide an exciting learning experience for

the students from AEARU member universities, connecting them with their peers from

around the world.

The Camp will be hosted by Peking University in Beijing, China. Courses will be focusing

on the frontier knowledge of Computer Science and Engineering. Both undergraduates

and graduate students are welcome to apply through the AEARU coordinator of their home

universities.

Courses:

Course Name Instructor Dates Credit

Deep Learning TANG Jian July 3 - 8 2 credits

Computational Social Science ZHU Jianhua Winson PENG July 2 - 12 2 credits

Probabilistic Models for Structured Data SUN Yizhou July 2 - 13 2 credits

Machine Learning in Computer Vision Carlo TOMASI July 2 - 15 2 credits

Foundations of Big Data Systems Tamer OZSU July 9 – 15 2 Credits

Biometric Authentication: System and Application

ZHANG Dapeng July 9 – 22 2 Credits

Health Informatics — Big Data Approach ZHANG Yanchun July 14 – 20 2 credits

Compact Data Structures for Big Data CHEN Shigang July 15 – 22 2 credits

Economics and Computation XIA Lirong July 16 – 25 2 Credits

Design Informatics Maria Wolters July 16-27 2 Credits

Becoming a Medtech Entrepreneur -- What is Biodesign?

Robert Chang July 16-27 2 Credits

1

https://sites.google.com/site/pkujiantang/

http://weblab.com.cityu.edu.hk/blog/jzhu

https://comartsci.msu.edu/our-people/taiquan-winson-peng

http://web.cs.ucla.edu/%7Eyzsun/cv/cv_yizhou_Jul_2017.pdf

https://users.cs.duke.edu/%7Etomasi/

https://cs.uwaterloo.ca/%7Etozsu/

http://www4.comp.polyu.edu.hk/%7Ecsdzhang/

https://www.vu.edu.au/contact-us/yanchun-zhang

http://www.cise.ufl.edu/%7Esgchen/

http://www.cs.rpi.edu/%7Exial/

http://mariawolters.net/


https://profiles.stanford.edu/robert-chang

Machine Learning for Time Series Analysis – Statistical Models and Deep Learning

LIU Yan July 23 – 27 2 Credits

Data Management for Big Data Analytics Leonid LIBKIN July 23-31 2 Credits

Computer Ethics Stephen COOPER LU Junlin July 23 – August 3 2 credits

Computation, Economics and Data Science CAI Yang July 23- August 3 2 Credits

Foundations of Artificial Intelligence Vincent NG July 25- August 5 2 Credits

Computational Game Theory Dan Garcia July 31-Aug 10 2 Credits

Attached please find the detailed syllabi of each course.

Eligibility:

Be between the ages of 18 and 40 (inclusive) and be in good health;

Be currently a full-time registered student at one of the AEARU member universities;

Have proof of English proficiency (absolved if the teaching language of the home

university is English).

Minimum GPA requirement of 3.0

Tuitions and Fees:

The tuition fee is WAIVED for nominated AEARU students.

Travel, accommodation, meals, and other expenses are the responsibility of the students.

For each AEARU member university, PKU can offer up to TWO places of FREE

accommodation for the period from July 15 to July 28, 2018 (based on the nomination of

the home university).

Note:

Sometimes, course(s) might be cancelled if the number of enrolled students is below the

minimum requirement. In such case, the affected student(s) can choose another course or

withdraw from the program. The confirmed list of courses to be provided before May 31,

2018.

2

http://www-bcf.usc.edu/%7Eliu32/

http://homepages.inf.ed.ac.uk/libkin/

https://directory.unl.edu/people/scooper22

http://cs.mcgill.ca/%7Ecai/

http://www.hlt.utdallas.edu/%7Evince/

http://www.cs.berkeley.edu/%7Eddgarcia/

Course Syllabus:

Course Name Page

Deep Learning 4

Computational Social Science 9

Probabilistic Models for Structured Data 15

Machine Learning in Computer Vision 20

Foundations of Big Data Systems 27

Biometric Authentication: System and Application 31

Health Informatics — Big Data Approach 33

Compact Data Structures for Big Data 38

Economics and Computation 47

Design Informatics 53

Becoming a Medtech Entrepreneur -- What is Biodesign? 62

Machine Learning for Time Series Analysis – Statistical Models and Deep Learning 69

Data Management for Big Data Analytics 73

Computer Ethics 81

Computation, Economics and Data Science 90

Foundations of Artificial Intelligence 96

Computational Game Theory 98

3

Peking University Summer School International 2018

Course Form for PKU Summer School International 2018

Course Title Deep Learning

深度学习

Teacher TANG Jian

First day of classes July 3, 2018

Last day of classes July 8, 2018

Course Credit 2 credits

Course Description

Objective

Deep learning has achieved tremendous success in many applications. This course aims to

introduce the fundamental concepts, methods, and applications of deep learning. The first

part of the course focuses on the theory and methods; the second part provides an

introduction to the widely used deep learning framework TensorFlow; the last part

introduces the applications of deep learning to various domains including computer vision,

natural language understanding, information network analysis, and recommendation.

Pre-requisites /Target audience

Data Structures and Algorithm, Probability and Mathematical Statistics.

Undergraduate students.

Proceeding of the Course

None

Assignments (essay or other forms)

Reading, Assignment and Programming

Evaluation Details

Attendance and Reading: 30%

Programming Project: 40%

Presentation: 30%

Text Books and Reading Materials

1. Ian Goodfellow, Yoshua Bengio, Aaron Courville. Deep Learning. MIT Press, 2016.

2. Christopher Bishop. Pattern Recognition and Machine Learning. Springer, 2013.

3. Trevor Hastie, Robert Tibshirani, Jerome Friedman. The Elements of Statistical

Learning. Springer, 2016.

4


4. Jure Leskovec, Anand Rajaraman, Jeffery D.Ullman. Mining of Massive Datasets.

Cambridge University Press, 2010.

5. Alex Smola and S.V.N. Vishwanathan. Introduction to Machine Learning.

Cambridge University Press, 20110.

6. Kevin P.Mruphy. Machine Lerning: A Probabilistic Perspective. MIT, 2012.

7. Mehryar Mohri, Afshin Rostamizadeh. Foundations of Machine Learning. MIT,

2012.

8. Tom M. Mitchell. Machine Learning. McGraw-Hill Education, 1997.

9. aser S. Abu-Mostafa, Malik Magdon-Ismail, Hsuan-Tien Lin. Learning from Data.

AML Book, 2012.

Academic Integrity (If necessary)

CLASS SCHEDULE

(Subject to adjustment)

Session 1： Math and Machine Learning Basics Date:7/2

【Description of the Session】(purpose, requirements, class and presentations

scheduling, etc.)

Introduce math and machine learning basics, such as probabilistic theory and matrix

computation.

【Questions】

【Readings, Websites or Video Clips】

【Assignments for this session (if any)】

Session 2：Feedforward Neural Networks Date:7/2


scheduling, etc.)

Introduce history of deep learning and some simple neural network structures.

【Questions】

5



Session 3：Optimization and Tricks Date:7/3


scheduling, etc.)

Introduce optimization methods in deep learning and effective tricks.

【Questions】



Session 4：Convolutional Neural Networks Date:7/3


scheduling, etc.)

Introduce convolutional neural networks and its applications.

【Questions】



Session 5：Recurrent Neural Networks Date:7/4

6



scheduling, etc.)

Some basics of RNN and corresponding applications in NLP.

【Questions】



Session 6：Introduction to TensorFlow Date:7/4


scheduling, etc.)

Introduce how to use TensorFlow.

【Questions】



Session 7：Deep Learning for Natural Language Understanding Date:7/5


scheduling, etc.)

Use deep learning solve NLP tasks.

【Questions】


7



Session 8：Deep learning for network analysis Date:7/5


scheduling, etc.)

Application of deep learning in network embedding and network data mining.

【Questions】



Session 9：Deep learing for recommendation and Student Presentation Date:7/6


scheduling, etc.)

Application of deep learning in recommendation task.

【Questions】

【Readings, Websites or Video Clips】 Yi Lu, Andrea Montanari, Balaji Probhakar,

Sarang Dharmapurikar, and Abdul Kabbani. Counter Braids: A Novel Counter

Architecture for Per-Flow Measurement. ACM SIGMETRICS, 2008.


8



Course Title Computational Social Science

计算社会科学

Teacher Jonathan ZHU, Winson PENG




Course Description

Objective

The course aims to introduce computational social science (CSS) to students in various data

science disciplines such as computer science, mathematics, statistics, electronic engineering,

bio-/medical informatics, etc., who are interested in individual human behaviors and

aggregate social processes. Assuming the enrolled students to have already known how to

perform basic data analysis using a variety of algorithms or tools, the course focuses on the

fundamental principles (e.g., what are good or bad practices) and the research design (e.g.,

how to plan rigorous studies to answer causal questions). Ample examples from the existing

literature will be used to help illustrate the principles and apply the research design. At the

end of the course, the students are expected to be able to do the following: (1)

communicating and collaborating with social scientists using their vocabulary and reasoning

logic; (2) planning and implementing rigorous computational studies on human behaviors or

social processes; (3) evaluating the quality of other social science studies and suggesting

feasible improvements.


Pre-requisites: Basic knowledge of statistical analysis, text mining, machine learning

Target audience: Senior undergraduate students and graduate students in various disciplines

of data science (computer science, mathematics, statistics, electronic engineering,

bio-/medical informatics, etc.)


No


Readings, In-class and online discussions, and take-home exercises

Evaluation Details

9


Session 1: Introduction to CSS and Research Design Date: 7/2/2018

【Description of the Session】(purpose, requirements, class and presentations scheduling,

etc.)

The basic features of computational social science (CSS, e.g., what it is, how it has come

about and why, what it aims to do, and how it works), primary purpose of social science

research (descriptive, explanatory, or predictive), and most common research designs in

social research (within-subjects design and between-subjects design).

【Questions】

What is CSS? How is it related to and different from social sciences and data science and

why? How to study social phenomena cross time, space, and population?


1. Babbie Ch1, Ch3-4

2. Salganik Ch1, Ch3

3. Peng & Zhu (2013)

4. Campbell Video1 1.1-1.7; 4.1-4.7; 5.1-5.8; 6.1-6.8


Nil

Session 2: Experiment in Social Research Date: 7/3/2018


etc.)

Principles and procedure of social science experiment (i.e., comparison between subjects

with and without exposure to stimulus) as the key solution to problems and challenges in

causal reasoning.

【Questions】

Why is it difficult to determine causes of social processes and how to deal with the

problems?

Attendance and Discussions: 25%

Assignments: 45%

Exam: 30%


Matthew J. Salganik (2018). Bit by Bit: Social Research in the Digital Age. Princeton

University Press;

Earl Babbie (2011). Principles of Social Research, 11th Edition. Tsinghua University

Press (or later edition after 11th from Wordsworth);

Additional readings are given in the Class Schedule.


Students are allowed to discuss readings and assignments among classmates in and outside

the class, but are discouraged to seek help from any living person outside the class.

However, individual-based writing assignments must be independently completed (i.e.,

without any plagiarism).

CLASS SCHEDULE


10

https://www.coursera.org/learn/social-science-study-chinese-society/lecture/spsYW/course-overview

https://www.coursera.org/learn/social-science-study-chinese-society/lecture/XTdQn/4-1-overview

https://www.coursera.org/learn/social-science-study-chinese-society/lecture/LaRYL/5-1-overview

https://www.coursera.org/learn/social-science-study-chinese-society/lecture/yNsP9/6-1-overview



1. Babbie Ch8

2. Salganik Ch4

3. Bond et al. (2012)

4. King et al. (2017)

5. Campbell Video1 7.1-7.8


Exercise 1. Review and critique of the research design and the causal reasoning used in a set

of published CSS studies

Session 3: Data for CSS Date: 7/5/2018


etc.)

How to collect data on human behaviors through online and offline means; technical,

operational, and quality challenges involved.

【Questions】

What are the major advantages and shortcomings of online data as compared with offline

data?


1. Babbie Ch9, Ch11

2. Salganik Ch2, Ch5

3. Liang & Zhu (2017)



Nil

Session 4: Sampling Date: 7/6/2018


etc.)

Principles and methods of classic probability sampling; opportunities and challenges of

applying probability sampling to online data.

【Questions】

When does probability sampling work for online data and when doesn’t? How to deal with

the problems?


1. Babbie Ch7

2. Zhu et al. (2011)

3. Xu & Zhu (2016)



Nil.

Session 5: Measurement Date: 7/7/2018


etc.)

How to quantify human thoughts and behaviors from empirical data; how to link latent

11

https://www.coursera.org/learn/social-science-study-chinese-society/lecture/0WxUf/7-1-overview

https://www.coursera.org/learn/social-science-research-chinese-society/lecture/qwtu7/2-1-evidence

https://www.coursera.org/learn/social-science-research-chinese-society/lecture/j7XAn/3-1-overview-sampling


concepts and observed variables; how to control for measurement noises.

【Questions】

How to evaluate and improve the quality (i.e., validity and reliability) of quantified data?


1. Babbie Ch5-6

2. Lazer et al. (2014)

3. Newman (2017)


Exercise 2. Review and critique of the data collection, sampling, and measurement used in a

set of published CSS studies.

Session 6: Multivariate Analysis Date: 7/9/2018


etc.)

Principles, procedure, and exemplar applications of multivariate analysis (e.g., multiple

regressions) to describe, explain, and predict causal processes with multiple causes.

【Questions】

How to control for confounding effects, identify inflated or suppressed effects, and

mediated/moderated effects?


1. Babbie Ch14-16

2. To be added


Nil.

Session 7: Multilevel and Temporal Analysis Date: 7/10/2018


etc.)

Multilevel analysis: principles, procedure, and exemplar applications of multilevel analysis

(e.g., mixed effects model and network analysis) to identify socially structured effects on

individuals and emergent effects from individuals on social structure. Temporal analysis:

principles, procedure, and exemplar applications of temporal analysis (e.g., time series

analysis, cohort analysis, survival analysis, sequence analysis) to model dynamic

characteristics of individual human behaviors and aggregated social processes.

【Questions】

Multilevel analysis: How to quantify individual variability within global regularities? How

to determine societal and collective effects on individuals? How to determine individual

contributions to the society? Temporal analysis: How to separate and quantify different

dynamic processes (e.g., auto-regression, periodicities, long-term trends, short-term shocks,

individual habits, social constraints, etc.) underlying social data?


1. Akcland & Zhu (2015)

2. Peng & Zhu (2012)

3. Wimmer & Lewis (2010)

4. Murdock et al. (2017)

12


5. Robinson & Deng (2015)

6. Xu et al. (2013)

7. Zhu et al. (2018)


Nil.

Session 8: Research Ethics and Student Presentation Date: 7/11/2018


etc.)

Ethical concerns for traditional social science research; new challenges for computational

social science; causes of and possible solutions to threats to privacy, confidentiality,

security, copyrights, and other issues. Students present the results of assignment 3.

【Questions】

When does computational social science research become harmful to individuals and the

society and how to mitigate the negative consequences?


1. Babbie Ch3

2. Salganik, Ch6-7

3. Campbell (2) Video 6.1-6.9


Exercise 3: Analysis and interpretation of online data with a complex structure (i.e.,

multivariate, multilevel, and time-stamped)

Sources of Textbooks, Videos, and Additional Readings

Textbooks/Videos:

Babbie = Earl Babbie (2007), The practice of social research, 11th ed. Tsinghua

University Press.

Salganik = Matthew J. Salganik (2018). Bit by bit: Social research in the digital age.

Princeton University Press.

Campbell = Cameron Campbell: Social science approaches to the study of Chinese

society, part 1 and part 2, Coursera.com

Additional Readings:

Ackland, R., & Zhu, J. J. H. (2015). Social network analysis. In P. Halfpenny & R,

Procter (Eds.), Innovations in digital research methods (pp. 221-244). Sage

Publications.

Bond, R. M., Fariss, C. J., Jones, J. J., Kramer, A. D. I., Marlow, C., Settle, J. E., &

Fowler, J. H. (2012). A 61-million-person experiment in social influence and

political mobilization. Nature, 489, 295.

King, G., Schneer, B., & White, A. (2017). How the news media activate public

expression and influence national agendas. Science, 358(6364), 776.

Lazer, D., Kennedy, R., King, G., & Vespignani, A. (2014). The Parable of Google

Flu: Traps in Big Data Analysis. Science, 343(6176), 1203-1205.

Liang, H., & Zhu, J. J. H. (2017). Big data, collection of (social media, harvesting).

In J. Matthes, C. S. Davis, & R. F. Potter (Eds.), International Handbook of

13

https://www.coursera.org/learn/social-science-research-chinese-society/lecture/T7hrh/6-1-overview-of-research-and-professional-ethics

https://www.coursera.org/learn/social-science-study-chinese-society/lecture/0WxUf/7-1-overview

https://www.coursera.org/learn/social-science-research-chinese-society/lecture/T7hrh/6-1-overview-of-research-and-professional-ethics


Communication Methods, Wiley & Sons.

Murdock, J., Allen, C., & DeDeo, S. (2017). Exploration and exploitation of

Victorian science in Darwin’s reading notebooks. Cognition, 159, 117-126.

Newman, M. (2017). Measurement errors in network data. arXiv preprint

arXiv:1703.07376.

Peng, T. Q., Zhang, L., Zhong, Z. J., & Zhu, J. J. H. (2013). Mapping the landscape

of Internet studies: Text mining of social science journal articles, 2000-2009. New

Media & Society, 15(5), 644-664.

Peng, T. Q., & Zhu, J. J. H. (2012). Where you publish matters most: A multilevel

analysis of factors affecting citations of internet studies. Journal of the American

Society for Information Science and Technology, 63(9), 1789-1803.

Robinson, W. N., & Deng, T. (2015). Data Mining Behavioral Transitions in Open

Source Repositories. Paper presented at the 2015 48th Hawaii International

Conference on System Sciences.

Wimmer, A., & Lewis, K. (2010). Beyond and Below Racial Homophily: ERG

Models of a Friendship Network Documented on Facebook. American Journal of

Sociology, 116(2), 583-642.

Xu, P., Wu, Y., Wei, E., Peng, T.-Q., Liu, S., Zhu, J. J. H., & Qu, H. (2013). Visual

Analysis of Topic Competition on Social Media. IEEE Transactions on

Visualization and Computer Graphics, 19(12), 2012-2021.

Xu, X. K., & Zhu, J. J. H. (2016). Flexible sampling large-scale social networks by

self-adjustable random walk. Physica A, 463, 356-365.

Zhu, J. J. H., Mo, Q., Wang, F., & Lu, H. (2011). A random digit search (RDS)

method for sampling of blogs and other web content. Social Science Computer

Review, 29(3), 327-339.

Zhu, J. J. H., Chen, H. X., Peng, T. Q., Liu, X. F., & Dai, H. X. (2018). How to

measure sessions of mobile device use? Quantification, Evaluation, and

Applications. Mobile Media & Communication.

14



Course Title Probabilistic Models for Structured Data

结构化数据的概率模型

Teacher SUN Yizhou




Course Description

Objective

The course aims to introduce probabilistic models for structured data, where data points are

no longer independent with each other, such as sequential data and graph/network data. The

course will cover modeling, inference, and learning of state-of-the-art probabilistic models,

including Hidden Markov Model, Markov Random Field, Conditional Random Field, and

Factor Graph. Applications across different domains, such as text mining, medical domain,

and social network analysis. At the end of the course, the students are expected to be able to

do the following: (1) understanding the mathematical formulation of different probabilistic

models that work for structured data, including intuition and mathematical derivations and

proof; (2) apply these models to real-world applications; (3) potential of developing novel

models for structured data for publications.


Pre-requisites: basic knowledge in statistics and probability, linear algebra, optimization,

programming.


(computer science, statistics, economics, finance, electronic engineering, biology, physics)


No



Evaluation Details


Assignments: 45%

Exam: 30%

15


Session 1: Introduction to Probabilistic Models and Structured Data Date: 7/2/2018


etc.)

Review of basics of probability theory and statistics; introduction to probabilistic models

and structured data; MLE and MAP principles; applications.

【Questions】

What are probabilistic models? What are structured data? What will structure bring in to

probabilistic models? What would be the standard procedure involved in probabilistic

modeling? What will be the useful applications of such models? What are the principles of

inference and learning on such models?



Problems on basic stats.

Session 2: Probabilistic Models for Unstructured Data Date: 7/3/2018


etc.)

Introduction of two well-known probabilistic models: Naïve Bayes and Logistic Regression,

and discuss of their limitations.

【Questions】

What is Naïve Bayes? What is logistic regression? What are their limitations? What are

generative models and discriminative models? How to extend Naïve Beyes to

semi-supervised setting?



Theoretical questions;

Implementation of two algorithms and apply them on a text classification task.


Daphne Koller and Nir Friedman (2009). Probabilistic Graphical Models. The MIT

Press;

Charles Sutton and Andrew McCallum (2014). An Introduction to Conditional Random

Fields. Now Publishers.

http://deepdive.stanford.edu/inference




the class, but are discouraged to seek help from any living person outside the class.

However, individual-based writing assignments must be independently completed (i.e.,

without any plagiarism).

CLASS SCHEDULE


16

http://deepdive.stanford.edu/inference


Session 3: Warm up: Hidden Markov Models Date: 7/4/2018


etc.)

Introduce HMM, which is a well-known probabilistic model for sequential data. Introduce

the concepts of modeling, inference, and learning via HMM.

【Questions】

What is sequential data? What are the applications? What is HMM? What are the standard

modeling, inference and learning procedure of HMM?



Session 4: Markov Random Fields Date: 7/5/2018


etc.)

Introduce MRF, a more general undirected graphical model. Introduce a simple pairwise

MRF.

【Questions】

What are MRFs? What is Markov property? What is collective inference? What is Gibbs

distribution? How to construct a simple pairwise MRF?




Implementation of simple pairwise MRF, with the application of text classification.

Session 5: Gaussian Markov Random Fields Date: 7/6/2018


etc.)

Introduce another special case of MRF, where random variables can take numerical values.

【Questions】

What is Gaussian MRF? What is the modeling, inference and learning procedure involved?




Implementation.

Session 6: Hinge Loss Markov Random Fields Date: 7/9/2018


etc.)

Introduce another special case of MRF, where reasoning can be performed. An application

on medical inference will be shown.

【Questions】

17


How to model human knowledge encoded as logic rules and apply these rules in the

modeling of MRF?



Brain storming on possible applications.

Session 7: Conditional Random Fields Date: 7/10/2018


etc.)

Introduce a directed probabilistic graphical model, which is CRF; introduce a special case of

CRF, linear-chain CRF, and its application on named entity recognition (NER).

【Questions】

What are CRFs? What is the difference between CRFs and MRFs? What are the pros and

cons.




Implementation and apply it on text classification.

Session 8: Skip-Chain Conditional Random Field Date: 7/11/2018


etc.)

Introduce another CRF, and its application to relation extraction.

【Questions】

What is the limitation of linear-chain CRF? What is general CRF? How does skip-chain

overcome the limitation of linear-chain CRF? How can skip-chain be applied to relation

extraction?



Session 9: Factor Graph Date: 7/12/2018


etc.)

Introduce a general form of probabilistic model, and its inference algorithm, sum-product

algorithm.

【Questions】

What is factor graph? How to do inference on it? What is sum-product algorithm? What

are the relationship between factor graph and the previous two models.



18


Session 10: Student Presentation / Exam Date: 7/13/2018


etc.)

Student presentation of previous homeworks and exam.

【Questions】



19



Course Title Machine Learning in Computer Vision

计算机视觉中的机器学习

Teacher Carlo TOMASI




Course Description

Objective

Machine learning methods have been employed for a long time in computer vision, but their

nature and impact on the field has changed dramatically since 2012, with the introduction of

deep networks. This course compares samples of pre-2012 and post-2012 machine learning

techniques in the context of image recognition, person detection, human body modeling,

and optical flow estimation. After an introduction to basic methods of image processing,

including the analysis of optical flow, the course surveys decision trees, random forests, and

deformable-parts models as a sample of pre-2012 methods, and reviews their use for person

detection and human body tracking. Deep convolutional neural networks are then discussed

in some detail, including convolutional pose machines and the OpenPose body tracking

system. The course closes with a review of supervised and unsupervised deep learning

methods for optical flow estimation.


Senior undergraduate and graduate students with a good grasp of linear algebra, probability,

and multivariate calculus. Some light-weight programming may be required for some of the

assignments, in a language chosen by the students.


Ten 3-hour lecture sessions and a final 2-hour exam. Readings are a mix of class notes and

papers from the literature. Each session is divided into three one-hour sections with brief

breaks in-between.


Daily reading assignments and short homework assignments with exam-style problems.

20


Evaluation Details

Attendance 20%; homework 45%; exam 35%


CLASS SCHEDULE


Session 1：Overview and Image Processing Date: July 2


scheduling, etc.) Overall structure and contents of the course. Elements of image

processing, including convolution, filtering, image differentiation, and image pyramids.

【Questions】


Class notes


Read materials for sessions 1 and 2 (due in session 2)

Session 2：Optical Flow Date: July 3


scheduling, etc.)

Two-frame image motion as optical flow: Definitions and variational estimation methods

for small displacements. The Shift-Invariant Feature Transform (SIFT). Large Displacement

Optical Flow (LDOF).

【Questions】


Class notes

D. Lowe. Distinctive image features from scale-invariant key points. International

Journal of Computer Vision, 60(2):91-110, 2004.

T. Brox and J. Malik. Large displacement optical flow: Descriptor matching in

variational motion estimation. IEEE Conference on Computer Vision and Pattern

21


Recognition, 41-48, 2009.


Homework 1 (due in session 4)

Session 3：A Machine Learning Sampler Date: July 4


scheduling, etc.)

Basic concepts of machine learning, including decision trees and random forests as

examples of “classical” machine learning methods.

【Questions】


Class notes


Read materials for sessions 3 (due in session 4)

Session 4：Person Detection and Body Tracking Date: July 5


scheduling, etc.)

Methods for finding people in images, and determining their body poses: Histograms of

Oriented Gradients (HOG), pedestrian detection, deformable models of body parts.

【Questions】


Class notes

N. Dalal and B. Triggs. Histograms of oriented gradients for human detection.

IEEE Conference on Computer Vision and Pattern Recogntiion, 1:886-893, 2005.

J. Gall and V. Lempitsky. Class-specific Hough forests for object detection. IEEE

Conference on Computer vision and Pattern Recognition, 143-157, 2009.

P. Felzenszwalb and D. Huttenlocher. Pictorial structures for object recognition.

International Journal of Computer Vision, 1:55-79, 2005.



22


Session 5：Convolutional Neural Networks Date: July 6


scheduling, etc.)

The basic structure of convolutional neural networks. Stochastic gradient descent as a

method for training deep networks.

【Questions】


Class notes



Session 6：Training Convolutional Networks. AlexNet. Date: July 9


scheduling, etc.)

Back-propagation of gradients through a deep network, and the use of stochastic gradient

descent for training. A description of AlexNet, the first broadly successful deep network in

computer vision.

【Questions】


Class notes.

A. Krizhevsky, I. Sutskever, and G. E. Hinton. ImageNet classification with deep

convolutional neural networks. Advances in Neural Information Processing Systems,

25:1106-1114, 2012.



Session 7：The State of the Art in Detection and Body Tracking Date: July 10


scheduling, etc.)

Recent deep networks achieve better-than human performance in object detection tasks, and

very high performance for body tracking. Two such systems are reviewed, based on the

concepts of batch normalization and convolutional pose machines, respectively.

23


【Questions】


S. Ioffe and C. Szegedy. Batch normalization: accelerting deep network training

by reducing internal covariate shift. International Conference on Machine Learning,

448-456, 2015.

S-E. Wei ,V. Ramakrishna, T. Kanade, and Y. Sheikh. Convoutional pose

machines. IEEE Conference on Computer Vision and Pattern Recognition,

4724-2732, 2017.

Z. Cao, T. Simon, S-E. Wei, and Y. Sheikh. Real time multi-person 2D pose

estimation using part affinity fields. IEEE Conference on Computer Vision and

Pattern Recognition, 1302-1310, 2017.


Read materials for session 8 (due in session 8)

Session 8： Supervised Deep Networks for Optical Flow Date: July 11


scheduling, etc.)

Optical flow is a regression problem, and contracting-expanding networks have proven up

to this task. Foundational concepts are reviewed, and a state-of-the-art convolutional flow

estimation method is studied.

【Questions】


A. Dosovitskiy, P. Fischer, E. Ilg, P. Hausser, C. Hazirbas, V. Golkov, P. van der Smagt, D.

Cremers, and T. Brox. FlowNet: Learning optical flow with convolutional networks.

International Conference on Computer Vision, 2758-2766, 2015.



Session 9：Unsupervised Optical Flow Estimation Date: July 12


scheduling, etc.)

Training an optical flow estimator requires large amounts of data. Synthetic datasets and

data augmentation are two techniques used to address this difficulty. Another one is to learn

flow in an unsupervised manner. Tradeoffs and techniques are studied in this session,

including spatial transformer networks and an unsupervised flow estimation system.

24


【Questions】


Class notes

M. Jaderberg, K. Simonyan, A. Zisserman, and K. Kavukcuoglu. Spatial

transformer networks. Advances in Neural Information Processing Systems,

2017-2025, 2015.

Z. Ren, J. Yan, B. Ni, B. Liu, X. Yang, and H. Zha. Unsupervised deep learning

for optical flow estimation. AAAI Conference on Artificial Intelligence, 1495-1501,

2017.


Read materials for session 9 (due in session 10)

Session 10：Review and Concluding Remarks Date: July 13


scheduling, etc.)

This section reviews what was covered in this course and what was not. The future of

machine learning and computer vision is discussed. Questions are answered in preparation

for the final exam.

【Questions】



Prepare for the final exam.

Session 11： Final Exam Date: July 15


scheduling, etc.)

The final exam takes two hours. It is closed book, closed notes, and covers the main

definitions and concepts through questions and simple problems.

【Questions】

25




26



Course Title Foundations of Big Data Systems

大数据系统基础

Teacher Prof. M. Tamer Özsu, University of Waterloo, Canada




Course Description

Objective：

The course addresses the foundations of modern big data systems. The focus is on data management

infrastructure. The course will address the fundamental challenges and components of big data

systems and approaches that have been developed to address them. The objective is that by the end

of this course, students should have a good understanding of the foundations of these systems.


Pre-requisites: database, data structure and algorithm

Target audience：senior undergraduate students, Master and PhD students


The course will review the foundational issues of big data systems. The focus is on data

management infrastructure. This infrastructure is typically built on top of modern distributed/parallel

computing platforms (e.g., MapReduce, Spark), run a distributed/parallel data management

platform, employ main memory systems (both row stores for OLTP and column stores for

analytics), and consist of multi-modal systems to handle different types of data coming from

different data sources. This course will cover these foundational issues.

The topics that will be covered are the following:

Fundamentals of distributed and parallel data management, focusing on data fragmentation,

distributed query processing, distributed transactions, replication, and data integration;

Main memory systems and column-based data representation;

Big data analytics platforms (distributed storage systems, MapReduce, Spark, graph

analytics, stream data management);

NoSQL, NewSQL and Polystore Systems (Key-Value Stores, Document Stores, Graph

Databases)

The course will consist of 20 hours of lectures, 4 hours per day , followed by students’

presentations (8 hours).


27


Evaluation Details

Students’ presentation.


There is no textbook that covers everything, but students will be provided with preprints of

chapters from the upcoming fourth edition of

M. Tamer Özsu and Patrick Valduriez, Principles of Distributed Database Systems,

Springer, forthcoming

In addition, there are some additional reading materials that are identified for each session

below.

Slides for all the classes will be provided on the course webpage.


CLASS SCHEDULE


Session 1: Fundamentals of distributed and parallel data management Date: 9 July 2018


scheduling, etc.) This session will cover the classical distributed/parallel data management

topics such as data partitioning and distribution, distributed query processing, distributed

transaction processing

【Questions】


Chapters 2, 4, 5, 6 of the forthcoming 4th edition of “Principles of Distributed Database

Systems”. These chapters will be provided on the course web page


Session 2： Main memory systems and column-based data representation Date: 10 July 2018

28



scheduling, etc.) An important aspect of big data systems is main memory processing and

column-based storage for data analytics. This session will cover these issues

【Questions】


F. Faerber, A. Kemper. P.A. Larson, J. evandoski, T. Neumann, and A. Pavlo, “Main

Memory Database Systems”, Foundations and Trends in Databases, 8(1-2): 1-130, 2016.


Session 3： Big data analytics platforms Date:10 July 2018


scheduling, etc.) This session will focus on the platforms that have been developed for big

data analytics. The specific topics that will be considered are MapReduce, Spark, and

stream processing.

【Questions】


Chapter 11 of the forthcoming 4th edition of “Principles of Distributed Database Systems”.

This chapter will be provided on the course web page.

Charu C. Aggarwal, Data Streams – Models and Algorithms, Springer, 2007.


Session 4： Graph analytics & graph databases Date: 11 July 2018


scheduling, etc.) Graphs have emerged as an important representation in big data systems.

In this session we consider both the graph analytics and graph database issues.

【Questions】

29



Chapters 11 and 12 of the forthcoming 4th edition of “Principles of Distributed Database

Systems”. These chapters will be provided on the course web page.

Da Yan, Yuanyuan Tian, and James Cheng, Systems for Big Graph Analytics, Springer,

2017.

Da Yan, Yingyi Bu, Yuanyuan Tian, and Amol Deshpande, Big Graph Analytics Platforms,

Foundations and Trends in Databases, 7(1-2): 1-195, 2017.


Session 5： NoSQL Systems Date: 12 July 2018


scheduling, etc.) We give an overview of NoSQL systems focusing on Key-Value Stores

and Document Stores

【Questions】


Chapter 12 of the forthcoming 4th edition of “Principles of Distributed Database Systems”.

This chapter will be provided on the course web page.


30

SUMMER SCHOOL IN PEKING UNIVERSITY SUBJECT DESCRIPTION

Subject title: Biometric Authentication: System and Application Subject code: _________________________________________________________________________________ Credit value: 2 credits Pre-requisite: (Subject title and code no, if any) Nil Recommended background knowledge: Nil Mutual exclusions: Nil Learning approach:

Lecture 28 hours Assessment:

Examination 100% Objectives: To understand the problems with current security systems.

To introduce biometric computing knowledge and methods.

To learn some basic biometrics systems based on the learned techniques.

Keyword syllabus: PART I: OVERVIEW Introduction to Biometrics Authentication

What is biometrics authentication? Traditional methods for personal authentication. Some definitions of biometrics authentication technologies and systems. Software and hardware biometrics systems.

Image processing and pattern recognition in living body, including human head & face, the mechanism of human eye, hand & skin characteristics. Biometrics Sensors and Data Acquisition

Biometric data acquisition and database. How to design various biometric sensors and how to evaluate their system performance? PART II: BIOMETRICS TECHNOLOGIES Biometrics Pre-processing

The related biometrics preprocessing technologies, including: noise removing, edge sharpening, image restoration, image segmentation, pattern extraction and classification. etc. Biometrics Feature Extraction

31

Basic elements in pattern recognition system, and some basic introduction of pattern recognition systems on biometrics (such as fingerprint, palm-print, finger, hand, face, iris, and face, as well as dental, DNA, and retina recognition). Features Matching and Decision Making

Various matching methods, including PCA and LDA. Introduce decision theory and their examples. PART III: TRADITIONAL BIOMETRICS Design and Implementation of Biometric Systems

Basic approaches of automated biometrics identification and verification systems. Various performance comparison and their analysis for large population authentication, accuracy and reliability of authentication in an e-world. Various biometrics (Fingerprint, Face, Iris, Palmprint, Signature, Voice, etc.) and their fusion technologies are introduced in detail. PART IV: ADVANCED BIOMETRICS 3D Biometrics and Multispectral Biometrics

Biometrics Fusion and Palmprint Recognition

Biometric Authentication Applications

Except various security applications, including access control, immigration and naturalization, military identification, banking, etc., some new applications, like medical biometrics and aesthetics biometrics, are also explored.

Indicative reading list and references: _ B. Zhang, Q. Zhao and D. Zhang, 2018, Facial Multi-Characteristics and Applications, 411pp,

ISBN 978-981-3234-57-4, World Scientific/Higher Education Press. _ D. Zhang, G. Lu and L. Zhang, 2018, Advanced Biometrics, 335pp, ISBN 978-3-319-61544-8,

Springer, Singapore.

_ D. Zhang, Y. Xu and W. Zuo, 2016, Discriminative Learning in Biometrics, 266pp, ISBN 978-981-10-2055-1, Springer, Singapore.

_ D. Zhang, F. Chen and Y. Xu, 2016, Computer Models for Facial Beauty Analysis, 268pp, ISBN 978-3-319-32596-5, Springer, USA.

_ D. Zhang, Z. Guo and Y. Gong, 2015, Multispectral Biometrics- Systems & Applications, 229pp, ISBN 978-3-319-22484-8, Springer, UK.

_ D. Zhang, W. Zuo and N. Li, 2015, Medical Biometrics-Computerized TCM Data Analysis, 363pp, ISBN 978-7-04-042883-4, World Scientific/Higher Education Press.

_ D. Zhang and G. Lu 2014, 3D Biometrics- Technologies and Systems, 290pp, ISBN 978-1-4614-7400-5. Springer, USA.

_ D. Zhang, F. Song, Y. Xu and Z. Liang, 2008, Advanced Pattern Recognition Technologies with Applications to Biometrics, IGI Global, USA, 369pp, ISBN 978-1-60566.

_ D. Zhang, X. Jing and J. Yang, 2005, Biometric Images Discrimination (BID) Technologies, IRM Press, USA, 357pp, ISBN 1-59140-831-8.

_ D. Zhang, 2004, Palmprint Authentication, Kluwer Academic Publishers, USA, 241pp, ISBN 1-4020-8096-4.

_ D. Zhang, 2000, Automated Biometrics: Technologies & Systems, Kluwer Academic Publishers, USA, 331pp, ISBN 0-7923-7856-3.

32



Course Title Health Informatics: Big Data Approach

健康信息学——大数据方法

Teacher ZHANG Yanchun




Course Description

Objective：

Health informatics is a cross discipline combining current information technology with

health sciences. Health informatics aims to improve the diagnosis of disease, achieve

effective treatment of disease and predict disease through collecting, processing and

managing health information. How to use current computer science, database and

information communication technology to improve disease diagnosis, and provide better

support for clinical decision making has already become one of issues in current medical

community. The purpose of the course is to enable students to be familiar with related topics

of health informatics and master the current advanced technique in health informatics. The

main content of this course includes health informatics, big medical data, big data analytical

methods, Internet of things for healthcare, medical images processing and its application,

sensing data streams mining, deep learning for healthcare applications and current research

topics in health informatics. This course is appropriate for students and researchers in the

computer science or medicine domains for helping them to understand related topics and

apply big data approaches for analyzing big medical data.


Data structures and Algorithms, Probability and Mathematical Statistics, Machine Learning

/ CS or medical students and researchers


None


Exercises

Evaluation Details

33


Attendance(10%)

Question Answering(40%)

Special Report(50%)

Text Books

1. Siuly Siuly, Yan Li, Yanchun Zhang: EEG Signal Analysis and Classification -

Techniques and Applications. Health Information Science, Springer 2016.

2. Xiao-Xia Yin, Sillas Hadjiloucas, Yanchun Zhang, Pattern Classification of Medical

Images: Computer Aided Diagnosis, Springer; June 29, 2017.

Reading Materials

1. Yanchun Zhang, Guiqing Yao, Jing He, Lei Wang, Neil R. Smalheiser, Xiao-Xia

Yin.Health Information Science.Springer 2014, ISBN 978-3-319-06268-6, 2014

2. Christopher Bishop. Pattern recogonition and Machine learning. Springer, 2007

3. Jiawei Han Mic, Micheline Kamber. Translated by Ming Fan, Xiaofeng Meng. Data

Mining: Concepts and Techniques. China Machine Press, 2001.

4. Tom Wbite. Translated by Minqi Zhou. Hadoop Authoritative Guide. Tsinghua

University Press, 2008.

5. Jimeng Sun, Chandan K. Reddy. Big Data Analytics for Healthcare. ACM SIGKDD

international conference on Knowledge discovery and data mining, 2013

6. Karla Caballero Barajas, Ram Akella. Dynamically Modeling Patient’s Health State from

Electronic Medical Records: A Time Series Approach. ACM SIGKDD international

conference on Knowledge discovery and data mining, 2015.

7. Guangyan Huang, Yanchun Zhang. Online mining abnormal period patterns from

multiple medical sensor data streams.World Wide Web 17(4): 569-587 (2014)

8. Jian Ma. An Introduction to the Technology of the Internet of Things. China Machine

Press, 2011

9. Robert Tibshirani, Trevor Hastie, Jerome Friedman. Elements of Statistical Learning.

Electronic Industry Press, 2004.

10. Akgul, Ceyhun Burak, et al. Content-based image retrieval in radiology: current status

and future directions. Journal of Digital Imaging 24.2 (2011): 208-222

11. Peter Schulam, Fredrick Wigley, Suchi Saria. Clustering Longitudinal Clinical Marker

Trajectories from Electronic Health Data: Applications to Phenotyping and Endotype

Discovery. Association for the Advancement of Artificial Intelligence, 2015

12. Xiang Wang, David Sontag, Fei Wang. Unsupervised Learning of Disease Progression

Models. ACM SIGKDD international conference on Knowledge discovery and data

mining,2014.


34

http://book.douban.com/search/Christopher%20Bishop

http://book.douban.com/search/Robert%20Tibshirani

http://book.douban.com/search/Trevor%20Hastie

http://book.douban.com/search/Jerome%20Friedman


CLASS SCHEDULE


Session 1：Health Informatics: Big Data Approach Date: 14/07/2018


scheduling, etc.)

In this session, we mainly introduce the basic concepts of health informatics, big medical

data, data collection, data processing and data management,

【Questions】



Session 2：Data Mining Methods and Applications Date: 15/07/2018


scheduling, etc.)

In this session, we mainly introduce the basic concepts and methods in data mining, such as

association rule mining, decision tree, Bayes and corresponding algorithms such as Apriori,

ID3, SVM.

【Questions】



Session 3：Internet of thing for healthcare, sensing data streams mining

and early-warning of medical monitoring

Date: 16/07/2018


scheduling, etc.)

In this session, we mainly introduce the basic concepts and application of the Internet of

things for healthcare, mining sensor data streams and medical anomaly detection,

early-warning and prediction.

35


【Questions】



Session 4： Internet of Things for Health Care: RFID Technology and

its Application

Date: 17/07/2018


scheduling, etc.)

In this session, We mainly introduce the basic concepts of the Internet of Things, RFID

（Radio Frequency Identification,）technology, uncertain data processing and applications

of Internet of Things.

【Questions】



Session 5, Deep learning for healthcare applications， medical images

processing

Date: 18/07/2018


scheduling, etc.)

In this session, we introduce the basic concepts, the existing models and applications of

deep learning in healthcare and medical images processing and its applications.

【Questions】


36



Session 6： Current research topics in health informatics Date: 19/07/2018


scheduling, etc.)

In this session, we introduce current research hot research topics in health informatics.

【Questions】



Session 7： Project presentation Date: 20/07/2018


scheduling, etc.)

In this session, we introduce related projects and students also have opportunity to show

presentation based on related topics.

【Questions】



37



Course Title Compact Data Structures for Big Data

紧凑数据结构与大数据

Teacher CHEN Shigang




Course Description

Objective

This course covers compact data structures, algorithms, probabilistic methods, and

statistical tools for handling big data, particularly in the setting of high-speed networks.

There is hardly any other data set whose size can rival the big network data that flows on the

Internet. Analyzing this big data is extremely useful in improving network performance,

user experience and cybersecurity. However, storing such data for analysis is impossible.

With this background, the course offers a series of compact data structures and their

theoretical analysis that are developed over the last three decades, with increasing

capabilities of making big data small for storage and quick access of data properties. The

offered data structures and their associated algorithms are broadly classified into two

categories: (1) counting and cardinality algorithms, including probabilistic counting, bitmap

algorithms, FM sketch, hyperloglog sketch, virtual bitmap, virutal FM sketch, virtual

hyperloglog, countMin, counter braids, randomized counter sharing, and virtual counters,

and (2) membership lookup and classification, including Bloom filters, counting Bloom

filters, Bloomier filters, blocked Bloom filters, multi-set filters, and multi-hashing tables.

The students will be exposed to not only these data structures and algorithms, but also their

applications in Internet traffic measurement, cybersecurity, as well as applications beyond

the network context. The students are expected to learn the compact data structures and

algorithms through lectures, and implement a selected subset with experiments over real

network data.


Data Structures and Algorithm, Computer Networks

Senior undergraduate students and graduate students


Data structure


38



Evaluation Details



Exam: 30%


1. Shigang Chen, Min Chen, Qingjun Xiao. Traffic Measurement for Big Network

Data. Springer, ISBN 978-3-319-47340-6, 2016.

2. Tao Li, Shigang Chen. Traffic Measurement on the Internet. Springer, ISBN:

978-1-4614-4850-1, 2012.

3. Kyu-young Whang, Brad T. Vander-Zanden, Howard M. Taylor. A Linear-Time

Probabilistic Counting Algorithm for Database Applications. ACM Transactions on

Database Systems, Vol. 15, No. 2, 1990.

4. Cristian Estan, George Varghese, Mike Fisk. Bitmap Algorithms for Counting

Active Flows on High Speed Links. ACM Internet Measurement Conference, 2003.

5. Peter Lieven, Bj orn Scheuermann. High-Speed Per-Flow Traffic Measurement with

Probabilistic Multiplicity Counting. IEEE INFOCOM, 2010.

6. Philippe Flajolet, Éric Fusy, Olivier Gandouet and Frédéric Meunier. HyperLogLog:

the analysis of a near-optimal cardinality estimation algorithm. Conference on

Analysis of Algorithms, 2007.

7. Qingjun Xiao, Shigang Chen, Min Chen, Yibei Ling. Hyper-Compact Virtual

Estimators for Big Network Data Based on Register Sharing. ACM SIGMETRICS,

2015.

8. Yi Lu, Andrea Montanari, Balaji Probhakar, Sarang Dharmapurikar, and Abdul

Kabbani. Counter Braids: A Novel Counter Architecture for Per-Flow Measurement.

ACM SIGMETRICS, 2008.

9. Andrei Brodery, Michael Mitzenmacherz. Network Applications of Bloom Filters: A

Survey, Internet mathematics, 2004.

10. Fang Hao, Murali Kodialam, T. V. Lakshman, and Haoyu Song. Fast Dynamic

Multiple-Set Membership Testing Using Combinatorial Bloom Filters. IEEE/ACM

TRANSACTIONS ON NETWORKING, VOL. 20, NO. 1, 2012.


CLASS SCHEDULE


Session 1： Probabilistic Counting Date:7/28

39



scheduling, etc.) basic methods and analytical tools for designing sketches

【Questions】

【Readings, Websites or Video Clips】 Kyu-young Whang, Brad T. Vander-Zanden,

Howard M. Taylor. A Linear-Time Probabilistic Counting Algorithm for Database

Applications. ACM Transactions on Database Systems, Vol. 15, No. 2, 1990.


Session 2：Bitmap Algorithms Date:7/29


scheduling, etc.) various bitmap algorithms for per-flow cardinality estimation

【Questions】

【Readings, Websites or Video Clips】 Cristian Estan, George Varghese, Mike Fisk.

Bitmap Algorithms for Counting Active Flows on High Speed Links. ACM Internet

Measurement Conference, 2003.


Session 3：FM Sketch Date:7/29


scheduling, etc.) operations and properties of FM sketches for per-flow cardinality

estimation

【Questions】

【Readings, Websites or Video Clips】 Peter Lieven, Bj orn Scheuermann.

High-Speed Per-Flow Traffic Measurement with Probabilistic Multiplicity Counting.

IEEE INFOCOM, 2010.

40



Session 4：Hyperloglog Sketch Date:7/30


scheduling, etc.) the state-of-the-art hyperloglog sketches for big data

【Questions】

【Readings, Websites or Video Clips】 Philippe Flajolet, Éric Fusy, Olivier Gandouet

and Frédéric Meunier. HyperLogLog: the analysis of a near-optimal cardinality

estimation algorithm. Conference on Analysis of Algorithms, 2007.


Session 5：Virtual Bitmap Date:7/30


scheduling, etc.) the basic methods and analytical tools for virtualized data structures

【Questions】



Session 6：Virtual FM Sketch Date:7/31


scheduling, etc.) different ways of virtualizing FM sketches

41


【Questions】



IEEE INFOCOM, 2010.


Session 7：Virtual Hyperloglog Date:7/31


scheduling, etc.) virtualizing hyperloglog

【Questions】

【Readings, Websites or Video Clips】 Qingjun Xiao, Shigang Chen, Min Chen, Yibei

Ling. Hyper-Compact Virtual Estimators for Big Network Data Based on Register

Sharing. ACM SIGMETRICS, 2015.


Session 8：CounterMin Date:8/1


scheduling, etc.) simple yet powerful countMin tools for summarizing big data

【Questions】



42


Session 9：Counter Braids Date:8/1


scheduling, etc.) extending countMin in two dimensions for better performance

【Questions】

【Readings, Websites or Video Clips】 Yi Lu, Andrea Montanari, Balaji Probhakar,

Sarang Dharmapurikar, and Abdul Kabbani. Counter Braids: A Novel Counter

Architecture for Per-Flow Measurement. ACM SIGMETRICS, 2008.


Session 10：Randomized Counter Sharing and Virtual Counters Date:8/2


scheduling, etc.) virtualizing counters

【Questions】



IEEE INFOCOM, 2010.


Session 11： Bloom Filters Date:8/2


scheduling, etc.) the basic ideas and analytical methods for Bloom filters

【Questions】

43


【Readings, Websites or Video Clips】 Andrei Brodery, Michael Mitzenmacherz.

Network Applications of Bloom Filters: A Survey, Internet mathematics, 2004.


Session 12： Counting Bloom Filters Date:8/3


scheduling, etc.) augmenting Bloom filters with the ability of deletion

【Questions】



Session 13：Bloomier Filters Date:8/3


scheduling, etc.) Augmenting Bloom filters with the ability of value lookup

【Questions】



Session 14：Blocked Bloom Filters Date:8/3

44



scheduling, etc.) A different design of Bloom filters

【Questions】

【Readings, Websites or Video Clips】 Andrei Brodery, Michael Mitzenmacherz.

Network Applications of Bloom Filters: A Survey, Internet mathematics, 2004.


Session 15：Multiset Bloom Filters Date:8/4


scheduling, etc.) variants of Bloom filters for multiset lookup

【Questions】

【Readings, Websites or Video Clips】 Fang Hao, Murali Kodialam, T. V. Lakshman,

and Haoyu Song. Fast Dynamic Multiple-Set Membership Testing Using

Combinatorial Bloom Filters. IEEE/ACM TRANSACTIONS ON NETWORKING,

VOL. 20, NO. 1, 2012.


Session 16： Multi-hash Tables Date:8/4


scheduling, etc.) using multiple hash functions to improve the space efficiency of hash

tables

【Questions】

45




46



Course Title Economics and Computation

经济与计算

Teacher XIA Lirong




Course Description

Objective

Economics and Computation is an emerging multi-disciplinary field of Economics,

Theoretical Computer Science, and Artificial Intelligence. It brings together principles and

methodologies in these fields to tackle challenges in the internet era. This course offers a

comprehensive in-depth introduction to key subjects in Economics and Computation. It will

cover great ideas in Economics, including key contributions of more than 10 Nobel

laureates in economics, as well as computational techniques and computational thinking in

new topics such as Algorithmic Game Theory and Computational Social Choice, which

were recognized one of the eleven “fundamental methods and application areas” of AI,

according to The One Hundred Year Study on Artificial Intelligence at Stanford University.

Students will learn (1) key applications of Economics and Computation, including social

choice, auctions, matching and resource allocation; (2) important conceptual contributions,

including Nash Equilibrium and their refinements, implementation theory, incentive

analysis, discrete choice models; (3) technical breakthroughs and algorithms, such as the

VCG mechanism, deferred acceptance algorithm, top-trading-cycles, generalized

method-of-moments; (4) modern topics such as security games, crowdsourcing, bitcoins.

This course is based on a highly-rated course taught by Lirong for four times at RPI.


The course only requires basic knowledge in algorithms. No background in Economics is

necessary.


None


47


Reading and written assignment

Evaluation Details


Written assignment: 40%

Exam: 30%


1. D. Parkes and S. Seuken. Economics and Computation (in progress).

2. L. Xia. Learning and Decision-Making from Rank Data (in progress).

3. F. Brandt, V. Coniter, U. Endriss, J. Lang, A. Procaccia. Handbook of Computational

Social Choice, 2016

4. Y. Shoham and K. Leyton-Brown, Multiagent Systems: Algorithmic,

Game-Theoretic, and Logical Foundations, 2009.

5. N. Nisan, T. Roughgarden, E. Tardos and V. Vazirani, Algorithmic Game Theory,

2007


Students are encouraged to work in groups, but they must acknowledge helps and

discussions with other students.

CLASS SCHEDULE


Session 1： Introduction to the course, basic game theory Date:7/16


scheduling, etc.)

Overview of the course, examples of EconCS systems. Basic game theory, pure Nash

Equilibrium. Mixed-strategy equilibrium, Nash's proof of existence.

【Questions】



Session 2： Sequential-Move games and Bayesian games Date:7/17

48

http://procaccia.info/papers/comsoc.pdf

http://procaccia.info/papers/comsoc.pdf

http://www.masfoundations.org/download.html

http://www.masfoundations.org/download.html

http://blog.oddhead.com/2007/09/17/computational-aspects-of-prediction-markets-book-chapter-and-extended-bibliography/



scheduling, etc.)

Extensive-form games, subgame perfect equilibrium. Bayesian games, Bayes-Nash

Equilibrium.

【Questions】



Session 3： Algorithmic Game Theory: Equilibrium Computation Date:7/18


scheduling, etc.)

Algorithms for computing mixed-strategy NE, correlated equilibrium, complexity of NE

【Questions】



Session 4： Mechanism design Date:7/19


scheduling, etc.)

Fundamental theory of mechanism design, implementation, revelation principle. Various

auction mechanisms, VCG mechanisms

【Questions】


49



Session 5： Matching and resource allocation Date:7/20


scheduling, etc.)

Deferred acceptance algorithm, top-trading-cycles algorithm, serial dictatorships.

Complexity and normative properties

【Questions】



Session 6： Voting Date:7/21


scheduling, etc.)

Definition and algorithms for voting rules. Axiomatic properties of voting rules, Arrow's

impossibility theorem.

【Questions】



Session 7： Computational Social Choice Date:7/22


scheduling, etc.)

Computational complexity and algorithms for winner determination. Using complexity to

protect elections. Strategic behavior and manipulation in social choice.

Gibbard-Satterthwaite theorem.

50


【Questions】



Session 8： Wisdom of the crowd Date:7/23


scheduling, etc.)

Condorcet Jury theorem and truthful elicitation, proper scoring rules.

【Questions】



Session 9： Discrete Choice Models Date:7/24


scheduling, etc.)

Discrete choice models, Plackett-Luce models, Random Utility Models, and generalized

method-of-moments algorithms for computing them.

【Questions】



51


Session 10： Crowdsourcing, Blockchain and bitcoin Date:7/25


scheduling, etc.)

Applications and challenges in crowdsourcing. Principles, protocols and strategic issues in

blockchain technology, illustrated in the context of bitcoin.

【Questions】



52



Course Title Design Informatics

设计信息学

Teacher Maria Wolters




Course Description

Objective

Most of the current technological advances are powered by data. The right data can

help us optimise devices, make services more efficient, and gain novel insights into

the world around us. Design Informatics is concerned with designing better

products, services, and technologies - for, with, and by data. We use Design

Thinking because it is particularly well suited for problems that do not have a single

correct solution - like most real world problems.

In this course, you will learn how to apply principles of User Centred Design to

create data-driven or data-centric solutions that really work for your users. In part 1

of the course, we review basic concepts and theories from design, human-computer

interaction, and human factors. In the second part, you will design a prototype for an

app, product, or service that provides data to users, collects data from users, or

adapts itself automatically to users.

Detailed readings and materials will be available on the course web site, which will

be announced on http://mariawolters.net and go live on June 1, 2018.


Computer science, Engineering, and Design students with an interest in Human

Computer Interaction.



53



User Requirements Analysis and User Research Project.


Readings, User Research Project, Design

Evaluation Details

Attendance, Readings: 30%

User Research Project: 40% (1000 words)

Final Design: 30% (short video)


1. Frank Ritter, Gordon Baxter, and Elizabeth Churchill (2014): Foundations

for Designing User-Centred Systems. Springer (FUCD)

2. http://interaction-design.org - Web site of the Interaction Design Foundation

(ID website)

3. http://usability.gov - Web site with basic guidance on measuring usability

(Usability web site)

4. http://inclusivedesigntoolkit.com - Web site on inclusive design (IDT

website)

Further readings will be provided on the course web page.


CLASS SCHEDULE


Session 1：Design Thinking I Date:7/16

【Description of the Session】Overview and general introduction to the

course

【Questions】

54

http://interaction-design.org/

http://usability.gov/

http://inclusivedesigntoolkit.com/


【Readings, Websites or Video Clips】selected materials from the ID

website


Session 2：Design Thinking II Date:7/16

【Description of the Session】Introduction to Design Thinking

【Questions】

【Readings, Websites or Video Clips】selected materials from the ID

website


Session 3: Introduction to Human Factors I Date:7/17

【Description of the Session】scope of Human Factors, perception,

cognition, anthropometry. Form and function in design.

【Questions】

【Readings, Websites or Video Clips】FUCD Chapters 1+2

【Assignments for this session (if any)】short commentary on readings

Session 4：Introduction to Human Factors II Date:7/17

55


【Description of the Session】Inclusive Design and accessibility.

Errors as data. Service Design.

【Questions】

【Readings, Websites or Video Clips】selected materials from the IDT

web site. FUCD Chapter 7


Session 5：Assessing User Requirements Date:7/18

【Description of the Session】What are user requirements? How can we

establish them? What are potential sources of data?

【Questions】


【Assignments for this session (if any)】short commentary on readings

Session 6：Personas and Scenarios Date:7/18

【Description of the Session】Personas and scenarios as tools for

describing user requirements. Agile user stories.

【Questions】

56


【Readings, Websites or Video Clips】Usability and ID web site -

selected resources


Session 7：Questionnaire Design Date:7/19

【Description of the Session】Introduction to designing and

administering questionnaires. Standardised Questionnaires

【Questions】

【Readings, Websites or Video Clips】selected readings from usability

and ID web site.

【Assignments for this session (if any)】completion and scoring of

sample questionnaires

Session 8：Questionnaire analysis Date:7/20

【Description of the Session】Questionnaire analysis, descriptive

statistics.

【Questions】



Session 9：Interviews Date: 7/23

57


【Description of the Session】Structured versus semi structured

interviews. Active Listening.

【Questions】


selected resources

【Assignments for this session (if any)】conduct and transcribe short

semi-structured interviews with classmates

Session 10：Interview Analysis Date:7/24

【Description of the Session】principles of qualitative analysis,

affinity diagramming

【Questions】


selected resources


Session 11：Initial Personas and Scenarios Date:7/24

【Description of the Session】Develop initial personas and scenarios

based on questionnaire and interview data

【Questions】

58




Session 12 User Observation Date:7/25

【Description of the Session】techniques for observing how people

interact with technology. Definitions of usability.

【Questions】

【Readings, Websites or Video Clips】selected materials from the

usability.gov website

【Assignments for this session (if any)】observe classmates according to

protocol developed in class

Session 13：Observation Analysis Date:7/26

【Description of the Session】Techniques for summarising user

observations

【Questions】




Session 14：Creating a Design Date:7/26

59




【Description of the Session】Group work on distilling findings from

user research into design

【Questions】



Session 15: User Journey Date: 7/27

【Description of the Session】Creation of a user journey based on the

design

【Questions】




Session 16：Summary Date: 7/27

【Description of the Session】Brief presentation of user journeys and

feedback

【Questions】

60





61



Course Title Becoming a Medtech Entrepreneur -- What is Biodesign?

生物医学技术创新与创业

Teacher Robert CHANG




Course Description

Objective

This course introduces the Stanford University Biodesign process

(http://biodesign.stanford.edu), which is design thinking for medical device innovation.

Also, the lean launch methodology will be integrated to cover customer development work

for commercialization. The overall course theme focuses on digital health, but concepts of

medical regulation, intellectual property, and startup fundraising are included. Thus, this

short course offers a broad view of medtech entrepreneurship to begin to demystify the

challenging path of healthcare innovation and is the basis for further study in unmet medical

needs, rapid prototyping, and business modeling. Students are expected to learn about the

vaunted Silicon Valley environment of innovation and to gain a new perspective on what it

takes to become a successful medtech entrepreneur. Ideally, small teams of students from

medicine, engineering, and business can apply to join the course together.


At least one developed skill focus in either medicine, engineering, or business

Senior undergraduate students and graduate students interested in healthcare innovation


Exposure to startups


Reading, Group Work, Final Presentation

Evaluation Details


Project: 40%

Exam: 30%

62



1. Yock, Zenios, Makower. Biodesign: the Process of Innovating Medical

Technologies. Cambridge University Press, ISBN-13: 978-1107087354, 2015.

2. Ries. The Lean Startup: How Today's Entrepreneurs Use Continuous Innovation to

Create Radically Successful Businesses. Crown Publishing Group, ISBN-13: 978-

0307887894, 2011.

3. Osterwalder. Business Model Generation: A Handbook for Visionaries, Game

Changers, and Challengers. John Wiley and Sons, ISBN-13: 978-0470876411, 2010.

4. Hou. Startups Demystified: Founders Share Strategies, Secrets, and Lessons

Learned. Coffee Cup Press, ISBN-13: 978-0692822951, 2018.

5. DiResta, Forrest, Vinyard. The Hardware Startup: Building Your Product, Business,

and Brand. O'Reilly Media, ISBN-13: 978-1449371036, 2015.

6. https://steveblank.com/books-for-startups/


CLASS SCHEDULE


Session 1： Biodesign Concepts, Digital Health, Strategic Focus Date:7/13


scheduling, etc.) Biodesign Overview and finding your niche.

【Questions】



Session 2：Need Statement, Observation Date:7/16

63

https://steveblank.com/books-for-startups/



scheduling, etc.) What is a need statement? How to identify unmet needs?

【Questions】



Session 3：Need Filtering, Specification Date:7/17


scheduling, etc.) Deciding on a top unmet need.

【Questions】



Session 4：Healthcare Overview Date:7/18


scheduling, etc.) Understanding how healthcare works.

【Questions】


64



Session 5：Design Thinking Principles, Brainstorming Date:7/19


scheduling, etc.) How to generate new ideas focused on a specific user.

【Questions】



Session 6：Business Model Canvas Date:7/20


scheduling, etc.) How to make it sustainable.

【Questions】



Session 7： Regulatory, Prototyping Date:7/23


scheduling, etc.) Typical medtech startup milestones.

65


【Questions】



Session 8： Intellectual Property, Incorporation, Equity Date:7/24


scheduling, etc.) What to do when you’re ready to start your company.

【Questions】



Session 9：Venture Capital, Fundraising Date:7/25


scheduling, etc.) Where does the money come from?

【Questions】



66


A CV of 250-300 words and a high-resolution personal photo should also be

provided

Title in English: Becoming a Medtech Entrepreneur -- What is Biodesign?

Professor：Robert Chang

Session 10：How to Pitch Date:7/26


scheduling, etc.) Crafting the ideal story.

【Questions】



Session 11： Final Pitches Date:7/27


scheduling, etc.) Final Project Presentations

【Questions】



67


Professor Chang is a medical innovator and an assistant professor of

ophthalmology on faculty at the Byers Eye Institute at Stanford University. As a full-

time glaucoma and cataract surgeon and clinician scientist, Dr. Chang spends his time

teaching the latest minimally invasive eye surgery techniques as well as researching

new technologies such as machine learning, portable imaging, wearables, digital

health, and telemedicine.

He is co-inventor of the first universal smartphone adapter to take pictures of the

front and back of the eye, licensed to Digisight Technologies as the FDA-cleared

PAXOS scope. He also co-founded a venture-backed connected health device startup

and has created an original extended hackathon medtech entrepreneurship course,

which has international exposure including the annual DreamCatchers MedTech

Hackathon in Hong Kong (www.dreamcatchers.hku.hk/), the Stanford Center at

Peking University in Beijing (www.dhealthclass.com/), and the HIPUC Bootcamp in

Brazil (hipuc.com, hilab.org).

Dr. Chang received his MD from the combined BA/MD program at the University

of Missouri, Kansas City School of Medicine. He completed his residency at the

prestigious Washington University in St. Louis, followed by a research and clinical

glaucoma fellowship at the renowned Bascom Palmer Eye Institute in Miami. He has

published over 40 peer-reviewed papers, delivered more than 100 invited national and

international lectures, been awarded many grants, holds multiple patents, and advises

medical device, biopharma, and startup companies on a regular basis. Currently, he is

Vice President of the Asia Pacific Tele-Ophthalmology Society (APTOS).

68

http://www.dreamcatchers.hku.hk/)

http://www.dhealthclass.com/)

http://hipuc.com/

http://hilab.org/



Course Title

Machine Learning for Time Series Analysis – Statistical

Models and Deep Learning

机器学习与时间序列分析

Teacher LIU Yan




Course Description

Objective

The course aims to introduce machine learning models, including both statistical models

and deep learning, to students in various science disciplines such as computer science,

statistics, economics, finance, electronic engineering, biology, physics etc., who are

interested in machine learning and statistical models for time series analysis and forecasting.

Enrolled students should have basic knowledge in statistics and probability, linear algebra,

optimization. No programming skills are required, but could be helpful for hands-on

exercise. The class covers popular time series models, including vector-regressive models

(VAR), ARIMA models, hidden Markov models, Karman filtering, as well as advanced

models, such as neural network models (Long Short-term memory neural networks,

recurrent neural networks, gated recurrent neural networks), support vector machine

regression, Hawkes processes, sparse VAR models etc. At the end of the course, the

students are expected to be able to do the following: (1) understanding the mathematical

formulation of time series models; (2) apply time series models to real-application data; (3)

potential of developing novel machine learning models for time series applications for

publications.


Pre-requisites: basic knowledge in statistics and probability, linear algebra, optimization.

No programming skills are required, but could be helpful for hands-on exercise.


(computer science, statistics, economics, finance, electronic engineering, biology, physics)


No


69


Session 1: Introduction to Time Series Date: 7/23/2018


etc.)

Definition of time series, stationary and non-stationary time series, white noise

Applications of time series analysis and forecasting

Introduction of basic time series models, moving average, auto-regression, ARMA models

and extensions

【Questions】

What is time series? What is stationary time series and non-stationary time series? What is

the basic models for time series analysis


1. Hamilton Ch 1-4, 11


Exercise 1. Questions on basic definitions of time series and stationary time series

Exercise 2. Hands-on exercise on running R-code for moving average, auto-regression,

ARMA models

Session 2: State-space models Date: 7/24/2018


etc.)

Introduction of state-space models, including hidden Markov model and Kalman filter

【Questions】

What is state-space models? What is hidden Markov model and Kalma filter? When should

we apply these models



Evaluation Details


Assignments: 45%

Exam: 30%


James D. Hamilton (1994). Time Series Analysis. Princeton Press;

Ian Goodfellow and Yoshua Bengio and Aaron Courville (2016). Deep Learning, MIT

Press.




the class. However, individual-based writing assignments must be independently completed

(i.e., without any plagiarism).

CLASS SCHEDULE


70


1. Hamilton Ch 5

2. Reading- Rabiner 1986


Exercise 3. Questions on hidden Markov models and Kalman filtering

Exercise 4. Hands-on exercise on hidden Markov models applying to text modeling

Session 3: Neural Network models for time series Date: 7/25/2018


etc.)

Introduction of basic neural network models, recurrent neural networks (RNN), long

short-term memory neural networks, gated recurrent neural networks

【Questions】

What is RNN? What type of properties does RNN capture in time series? How to apply

RNN for time series forecasting and prediction


1. Goodfellow et al, 2016 Chapter 4-6, 10


Exercise 5: Hands-on exercise on RNN for time series analysis

Session 4: Sparse VAR models and Granger causality Date: 7/26/2018


etc.)

Introduction of lasso, VAR models, sparse VAR models, Granger causality and extensions

【Questions】

How can we learn temporal dependencies? How can we address problems in practical

applications, such as nonstationary, irregular time series, relational time series?


1. Reading- Liu 2018


Exercise 6: Hands-on exercise on sparse-VAR models for time series dependence analysis

Session 5: Hawkes Process, and Support Vector Regression Date: 7/27/2018


etc.)

Introduction of Poisson process, Hawkes processes, support vector machines, and support

vector regression

【Questions】

How can we model stochastic data? How can we go beyond linear predictors?


1. Reading- Laub et al, 2015

2. Reading- Burges, 1998


Exercise 7. Hands-on exercise on Hawkes process models

71


Sources of Textbooks, Videos, and Additional Readings

Textbooks/Videos:

N/A

Additional Readings:

Rabiner. An introduction of Hidden Markov Model, 1986

Patrick J. Laub, Thomas Taimre, Philip K. Pollett. Hawkes Processes. 2015.

Yan Liu. A tutorial on sparse vector autoregression and granger causality, 2018.

Christopher J.C. Burges. A Tutorial on Support Vector Machines for Pattern

Recognition.

72



Course Title Data Management for Big Data Analytics

面向大数据分析的数据管理

Teacher Prof. Leonid LIBKIN, University of Edinburgh



Course Credit 2 credit

Course Description

Objective：

The course is about data management aspects of Big Data. Up to 80% of the big data effort is

what is commonly known as data wrangling – preparing data for machine learning and data

mining algorithms. In fact traditional databases remain the main tool of data analysts. The

course aims to introduce students to challenges of big data, and prepare them to conducting

research, in both academic and industrial settings, in the areas of querying and managing big

data, and expose them to current research and development in connection with big data theory.

This course will cover foundational issues in connection with three of four big V’s in the typical

characterization of big data, namely, Volume, Variety and Veracity.


Pre-requisites: database, data structure and algorithm, Discrete mathematics

Target audience：senior undergraduate students, Master and PhD students


The course will review fundamental challenges introduced by querying big data, such as

the need for revising the classical computational complexity theory in the context of big data.

Regarding Volume, it will deal with the feasibility of computing exact query answers in big

data within our available resources, and approximate query answering. For Variety, it will

cover popular data models, including relational, XML, graph, and RDF models, and languages

for them, as well as handling queries over data residing in multiple sources, focusing on both

virtual and materialized integration, and efficient query answering. For Veracity, it will cover

handling poor quality information, understanding current technologies and their deficiencies,

correctness guarantees, and consistent query answering, and will look into how ontologies help

produce better query answers. Students will be introduced to the study of several query

languages including SQL for relational databases, Cypher for graph data, and SPARQL for

RDF.

Specific topics will include:

Joins and conjunctive queries: evaluation and analysis

Scalable query answering

73


Approximation of queries

XML databases

Graph databases: path queries and patterns

Graph databases: querying property graphs

Querying RDF data

Incomplete information and correct query answering

Handling inconsistent data

Data integration

Data exchange

Ontology-mediated query answering

The course will consist of 28 hours of lectures, 4 hours per day , followed by students’

presentations.


Evaluation Details

Students’ presentation.


As the material is largely new, there is no single textbook that presents it. Some aspects are

covered in existing books, e.g. conjunctive queries in

1. Serge Abiteboul, Richard Hull, Victor Vianu, "Foundations of Databases", Addison-Wesley

Publishing Company, 1995

or data exchange in

2. Marcelo Arenas, Pablo Barceló, Leonid Libkin, Filip Murlak, “Foundations of Data

Exchange”, Cambridge University Press New York, NY, USA ©2014

Slides for all the classes will be provided on the course webpage.


CLASS SCHEDULE


Session 1: SQL as a data analytics tool Date:

74



scheduling, etc.) What is the most common tool used by data analysts? It is, as

multiple surveys show, SQL - the main query language for commercial

RDBMSs. We give a gentle reminder of what SQL is, and then point out some

serious issues that arise when one relies on SQL queries for data

analytics.

【Questions】



Session 2： Conjunctive queries: evaluation Date:


scheduling, etc.) Conjunctive queries, also known as select-project-join queries, are the

most fundamental queries used in database management systems. Since database

design principles prescribe splitting data into multiple tables, such joins need to be

taken to obtain useful information. Naïve evaluation of conjunctive queries however is

a computationally expensive problem. In this session we look at ways of speeding up

conjunctive query evaluation, and at their static analysis for efficient optimizations.

【Questions】



Session 3： Scalable query answering Date:


scheduling, etc.) These days we deal with enormous data repositories, so large that

even a linear time algorithm over them can take days or weeks. To answer queries, we

need new notions of complexity. The key idea comes in the form of scale-independence

(even if data is huge, the part relevant to the query is likely to be small). We study the

concept and look at access information that lets us find scalable queries.

【Questions】

75




Session 4： Approximation of queries Date:


scheduling, etc.) If, due to the size of data or complexity of the query, it is infeasible to

find exact query answer, we need to approximate query results. We look into

approximations of joins and conjunctive queries by queries with complexity

guarantees.

【Questions】



Session 5： XML databases Date:


scheduling, etc.) We give an overview of XML navigational languages, and connect

them with specification languages used in software and hardware verification. We also

employ the connection to look at static analysis of XML queries.

【Questions】



Session 6： Graph databases: path queries Date:

76



scheduling, etc.) Graph databases are becoming popular due to new applications such

as social networks and the Semantic Web. We look at models of graph data and

languages for them, based on path queries and graph patterns, and discuss the

complexity issues that arise in the evaluation of such queries.

【Questions】



Session 7: Graph databases: querying property graphs Date:

【Description of the Session】In most products such as those by Oracle, SAP, Neo4j,

the model that is predominantly used is of property graphs: in those, nodes and

relationship carry sets of key-value pairs that can be queries. We look at languages

that have been developed for such graphs, including both theoretical extensions of

path queries, as well as Cypher, a pattern based language of Neo4j, and analogs of

XPath for graphs.

【Questions】



Session 8： Querying RDF data Date:


scheduling, etc.) RDF is the formal underlying the Semantic Web; it is essentially a

form of graph data where labels and nodes can be mixed. We look at languages for

RDF, such as SPARQL, from database perspective and show that they have very

natural counterparts in the relational database world.

【Questions】

77




Session 9： Incomplete information Date:


scheduling, etc.) It is well known that standard relational languages such as SQL

produce very counter-intuitive results when information is incomplete. We study

formal models of correct answers to queries over incomplete data, and explain that

SQL, as is currently implemented, differs from them in all possible ways. We also

show how to fix query evaluation so that it would eliminate incorrect query answers.

【Questions】



Session 10： Inconsistent data and consistent query answering Date:


scheduling, etc.) Inconsistency arises when a database does not satisfy prescribed

specification; often this a byproduct of merging several databases. If data cannot be

cleaned, one needs to query inconsistent data. We introduce a model of such querying

and study its computational costs.

【Questions】



Session 11： Data integration Date:

78



scheduling, etc.) In data integration, data needs to be pulled from various sources and

queries. Often though it is infeasible to actually move data and restructure it under a

new schema, in which case the querying process is completely virtual. We look at

techniques for such virtual data integration based on query rewriting.

【Questions】



Session 12： Data exchange Date:


scheduling, etc.) Data exchange is the problem of moving data between applications.

These can rely on data structured according to different schemas, and thus one needs

schema mappings to reconcile them. We look at building target instances based on

schema mappings, answering queries over them, and analysis of metadata, i.e.,

mappings themselves.

【Questions】



Session 12： Ontology-mediated query answering Date:

【Description of the Session】Often data comes together with additional knowledge in

the shape of an ontology. Using such an ontology can improve the quality of query

answers. We look at some ontology languages that people use, and describe

algorithms that use ontologies to facilitate query answering.

【Questions】

79




80



Course Title Computer Ethics

计算机伦理学

Teacher Steve Cooper, Junlin LU


Last day of classes August 3, 2017


Course Description

Objective

Computer ethics is a very important topic for all students and professionals in

computer-related areas. This course focuses on computer ethics, including ethical and social

issues related to the development and use of computer technology. The main topics include

ethical theory, and social, political, and legal considerations. It covers scenarios in problem

areas: privacy, reliability and risks of complex systems, and responsibility of professionals

for applications and consequences of their work.


Undergraduate Students and Graduate Students



Reading, Research Project and Report

Evaluation Details

Attendance: 20%

Debate: 20%

Paper: 30%

Research Project: 30%


1. Computer Ethics: A Global Perspective, Giannis Stamatellos, Jones & Bartlett, 2007.

81


2. Computer Ethics (4th edition), Deborah G. Johnson, Pearson, 2009.

3. A Gift of Fire: Social, Legal, and Ethical Issues for Computing Technology (4 th edition),

Sara Baase, Pearson, 2012.

4. 火的礼物：人类与计算技术的终极博弈(第四版), Sara Baase (郭耀译), 电子工业出

版社, 2015.


CLASS SCHEDULE


Session 1：Course Introduction and Arrangement Date:7/23


scheduling, etc.) Introduction of the course and the arrangement for next ten days.

【Questions】



Session 2：Information Security Date:7/23


scheduling, etc.) Information security in today’s world and how to solve it.

【Questions】


82



Session 3：software reliability Date:7/23


scheduling, etc.) Are the software we are using in our daily lives safe? Give safety

analysis of software.

【Questions】



Session 4：Risks of Computer Date:7/24


scheduling, etc.) Give the analysis of computer-related risks.

【Questions】



Session 5：The Philosophical Basis of Ethics Date:7/24


scheduling, etc.) Describe the philosophical foundation of ethics. Introduce the concept

of ethics.

83


【Questions】



Session 6：Computer Ethics Date:7/25


scheduling, etc.) Introduce the relationship between computer and ethics.

【Questions】



Session 7：Computing and economics Date:7/25


scheduling, etc.) Introduce the relationship between computing and economy.

【Questions】



84


Session 8：Intellectual Property Date:7/26


scheduling, etc.) Introduce some relevant laws, Intellectual Property.

【Questions】



Session 9：Hacking Date: 7/26


scheduling, etc.) Introduce the Hacker Technology.

【Questions】



Session 10：Policy Problem：Data Leakage Date: 7/27


scheduling, etc.) There is a policy problem, which is Data Leakage.

【Questions】

85




Session 11：Monopoly Power Date: 7/27


scheduling, etc.) What is a monopoly and how it works.

【Questions】



Session 12：Internet History and Culture Date: 7/30


scheduling, etc.) Introduce the history and culture of the internet.

【Questions】



Session 13：Freedom of Speech on The Internet Date: 7/30

86



scheduling, etc.) The internet is open for everyone, is it completely freedom of speech

on it?

【Questions】



Session 14：Privacy Right Date: 7/31


scheduling, etc.) What is Privacy Right? What is invasion of privacy?

【Questions】



Session 15：Mathematical Gap Date: 7/31


scheduling, etc.) Introduce the Mathematical Gap.

【Questions】


87



Session 16：Computer and Education Date: 8/1


scheduling, etc.) Introduce the relationship between computer and education.

【Questions】



Session 16：Group Debate Date: 8/2


scheduling, etc.) Students participate in terms and debate in groups.

【Questions】



Session 16：Project Report Date: 8/3


scheduling, etc.) Students give a presentation to report their project.

88


【Questions】



89



Course Title Computation, Economics and Data Science

计算、经济与数据科学

Teacher CAI Yang




Course Description

Objective：

The course will present topics at the intersection of Algorithms, Game Theory, Economics,

and Learning. We will mainly focus on the following three topics: (i) Fundamentals of

Game Theory and their connection to duality theory and online learning; (ii) The basics and

foundations of auctions and mechanism design, and also how to design good auctions and

mechanisms using data through the lens of provably-approximately-correct (PAC) learning;

and (iii) The basics of density estimation and its applications in Econometrics.

Topic (i) will cover the basics of strategic behavior, equilibria, duality theory, online

learning, and the price of anarchy. Topic (ii) will present the basics of mechanism design,

revenue optimization, and PAC learning, and apply these tools to study the simplicity,

learnability, and approximation tradeoffs in mechanism design. Topic (iii) will present the

basics of Econometrics and Statistics, with applications to inference in games and auctions.

Most examples of this course will be based on applications in online advertising and online

market design.




90


Evaluation Details



CLASS SCHEDULE


Session 1：Title Date:


scheduling, etc.)

【Questions】





scheduling, etc.)

【Questions】

91






scheduling, etc.)

【Questions】





scheduling, etc.)

【Questions】




92



scheduling, etc.)

【Questions】





scheduling, etc.)

【Questions】





scheduling, etc.)

【Questions】


93





scheduling, etc.)

【Questions】





scheduling, etc.)

【Questions】





scheduling, etc.)

94


【Questions】



95

SUMMER SCHOOL IN PEKING UNIVERSITY SUBJECT DESCRIPTION

Subject title: Fundamentals of Artificial Intelligence Subject code: _________________________________________________________________________________ Credit value: 2 credits Pre-requisite: (Subject title and code no, if any) Nil Recommended background knowledge: Data Structures and Algorithm, Discrete Mathematics (logic, probability, counting) Mutual exclusions: Nil Learning approach:

Lecture 28 hours Assessment:

Examination 100% Objectives: This course introduces the theoretical and computational techniques that serve as a foundation for the study of artificial intelligence (AI). Specific objectives include: understanding basic search algorithms

understanding algorithms used for logical and probabilistic reasoning

acquiring the basics of game theory Keyword syllabus: PART I: SEARCH Uninformed Search

Problem solving as search, breadth-first search, depth-first search, uniform-cost search, iterative deepening, bidirectional search

Informed Search

Greedy best-first search, admissible heuristics, A*

Local Search

Hill-Climbing search, simulated annealing, Davis-Putnam, satisfiability, genetic algorithms

Constraint Satisfaction

Backtracking, value and variable selection heuristics, forward checking, constraint propagation,

problem encoding as CSPs

Adversarial Search

Game playing as search, simple minimax, heuristic minimax, alpha-beta pruning, expectiminimax 96

PART II: KNOWLEDGE REPRESENTATION AND REASONING Basics of Knowledge Representation and Reasoning

Knowledge representation, logic, soundness and completeness of proof theory

Propositional Logical Reasoning

Propositional logic, rules of inference, resolution, chaining

First-Order Logical Reasoning

First-order logic, rules of inference, resolution PART III: PROBABILISTIC REASONING Review of Probability Theory

Basics of probability theory, uncertainty, the joint probability distribution, conditional independence

Semantics of Bayesian Networks

Bayesian networks, d-separation

Exact Inference

Enumeration, variable elimination

Approximate Inference

Stochastic simulation, likelihood weighting PART IV: GAME THEORY Games with Hidden Information

Matrix normal form of games, pure and mixed strategies

Non-Zero Sum Games

Prisoner’s dilemma, Nash equilibrium, tragedy of the commons

Indicative reading list and references: _ S. Russell and P. Norvig, 2010. Artificial Intelligence (3rd edition). 1132pp, ISBN 978-0-13-

604259-4, Prentice Hall.

97



Course Title Computational Game Theory

计算对弈游戏理论

Teacher Dan Garcia


Last day of classes August 10, 2018


Course Description

Objective

This course will explore the fertile ground of computational game theory, i.e., the use of

computers to solve abstract strategy games via computational brute force. We blend ideas

from artificial intelligence, combinatorics, human-computer interface design, parallel and

distributed computing and software engineering. Students will learn the mathematical

foundations of abstract strategy games and the special case of combinatorial games,

including surreal numbers. They will learn how to calculate the upper bounds of the size of

a game tree, will author a simple recursive solver for loop-free games, and a retrograde

solver to handle loopy games. They will extend their retrograde solver to work in a

distributed environment using the Apache Spark™ API. For their final project, students will

work in teams to choose and encode an abstract strategy game, solve it, author a graphical

interface that will allow the system to play perfectly, and use it to perform analysis.


Data Structures and Algorithms, Discrete Mathematics, Python language fluency



Data structures



Evaluation Details

Attendance, Participation and Reading: 20%


Final Report: 30%

98


REQUIRED Text Books and Reading Materials

1. Jonathan Schaeffer, Neil Burch, Yngvi Björnsson, Akihiro Kishimoto, Martin

Müller,Robert Lake,Paul Lu, Steve Sutphen, Checkers Is Solved. Science 14 Sep

2007:Vol. 317, Issue 5844, pp. 1518-1522, DOI: 10.1126/science.1144079

2. R.U.Gasser. Solving Nine Men's Morris R.J.Nowakowski (Ed.), Games of No

Chance, MSRI Publications, 29,Cambridge University Press, Cambridge, MA

(1996), pp.101-113

3. H.Jaapvan den Herik, Jos W.H.M.Uiterwijk, Jackvan Rijswijck. Games solved:

Now and in the future. Artificial Intelligence, Volume 134, Issues 1–2, January

2002, Pages 277-311. https://doi.org/10.1016/S0004-3702(01)00152-7

RECOMMENDED Text Books and Reading Materials

1. Elwyn R. Berlekamp, John H. Conway, Richard K. Guy. Winning Ways for Your

Mathematical Plays: Volume 1 (2nd Edition), A K Peters/CRC Press;

978-1568811307; January 18, 2001)


CLASS SCHEDULE


Session 1： Introduction to Computational Game Theory Date:7/31


scheduling, etc.) After group introductions, we will provide an overview of the field,

motivation and basics of computational game theory. We will also explore many popular

abstract strategy games and form classifications.

【Questions】


1. H.Jaapvan den Herik, Jos W.H.M.Uiterwijk, Jackvan Rijswijck. Games solved:

Now and in the future. Artificial Intelligence, Volume 134, Issues 1–2, January

2002, Pages 277-311. https://doi.org/10.1016/S0004-3702(01)00152-7

【Assignments for this session (if any)】Author a recursive backtracking solver in Python

for the Nim-like game 10, 9, … 1 (starting with 10 coins, remove one or two on your turn

with the goal of removing the last coin) and rudimentary playing system.

Session 2：Determining the Size of Games Date:8/01

99

https://www.sciencedirect.com/science/journal/00043702

https://www.sciencedirect.com/science/journal/00043702/134/1

https://doi.org/10.1016/S0004-3702(01)00152-7

https://www.sciencedirect.com/science/journal/00043702

https://www.sciencedirect.com/science/journal/00043702/134/1

https://doi.org/10.1016/S0004-3702(01)00152-7



scheduling, etc.) Using principles from Combinatorics, students will learn how to calculate

bounds on the number of positions of games.

【Questions】

【Readings, Websites or Video Clips】 YouTube “Numberphile: Connect 4”

https://www.youtube.com/watch?v=yDWPi1pZ0Po

1. R.U.Gasser. Solving Nine Men's Morris R.J.Nowakowski (Ed.), Games of No

Chance, MSRI Publications, 29,Cambridge University Press, Cambridge, MA

(1996), pp.101-113

【Assignments for this session (if any)】Determine the number of positions in Connect 4

in closed form. Polish the “game player” from Session 1.

Session 3：Solvers Date:8/02


scheduling, etc.) Starting from a simple recursive descent solver, we will understand its

limitations, and will first increase its speed significantly through memoization, and then

explore the retrograde algorithm to handle loopy games, as well as explore an optimized

iterative solver for tier games.

【Questions】

【Readings, Websites or Video Clips】 https://en.wikipedia.org/wiki/Hare_games

【Assignments for this session (if any)】 Author a retrograde solver in Python and test it

on a small loopy game (e.g., Hare and Hounds, Chopsticks, etc)

Session 4：Distributed Solvers Date:8/03


scheduling, etc.) Students will be introduced to the Apache Spark system and explore how

to use it to speed up an iterative solver.

【Questions】

100

https://www.youtube.com/watch?v=yDWPi1pZ0Po



1. Jonathan Schaeffer, Neil Burch, Yngvi Björnsson, Akihiro Kishimoto, Martin

Müller,Robert Lake,Paul Lu, Steve Sutphen, Checkers Is Solved. Science 14 Sep

2007:Vol. 317, Issue 5844, pp. 1518-1522, DOI: 10.1126/science.1144079

【Assignments for this session (if any)】Students will choose a game to implement for

their final project and begin working on it.

Session 5： Introduction to Combinatorial Game Theory Date:8/07


scheduling, etc.) Starting with Nim and moving quickly through to Kayles and

Domineering, students will learn the fundamentals of combinatorial game theory.

【Questions】


【Assignments for this session (if any)】Practice with the algorithm for playing Nim

perfectly, and continue to work on their final project.

Session 6：Building a Graphical User Interface Date:8/08


scheduling, etc.) Students will be introduced to the principles behind building a graphical

user interface for their game.

【Questions】


【Assignments for this session (if any)】 Begin to add an a graphical user interface onto

their final project

Session 7：Group Project Work Session Date:8/09

101


A CV of 250-300 words and a high-resolution personal photo should also be

provided

Title in English: Computational Game Theory

Professor：Dan Garcia


scheduling, etc.) This session is provided to allow the student groups to finish up their

projects with instructor guidance, and do analysis of the games.

【Questions】


【Assignments for this session (if any)】Continue working on the final project

Session 8：Final Project Demonstrations Date:8/10


scheduling, etc.) This session is devoted to demonstrations of the student final projects and

submissions of their final report.

【Questions】



102


Dan Garcia ([email protected]) is a Teaching Professor in the Computer

Science Division of the EECS Department at the University of California, Berkeley,

and joined the faculty in the fall of 2000. Dan received his PhD and MS in Computer

Science from UC Berkeley in 2000 and 1995, and dual BS degrees in Computer

Science and Electrical Engineering from MIT in 1990. He was chosen as an ACM

Distinguished Educator in 2012. He won NCWIT's Undergraduate Research

Mentoring (URM) Award in 2016.

He has won all four of the department’s teaching awards (the Information Technology

Faculty Award for Excellence in Undergraduate Teaching in 2004, the Diane S.

McEntyre Award for Excellence in Teaching in 2002, the EECS outstanding graduate

student instructor award in 1998, and the CS outstanding graduate student instructor

award in 1992.). He was also chosen as a UC Berkeley “Unsung Hero” in 2005. He

holds the record for the highest teaching effectiveness ratings (6.7/7) in the history of

the department's lower-division introductory courses. His research interests are

computer science education and computational game theory.

103

mailto:[email protected]

the aearu stem summer camp program 2018...5. campbell video1 7.1-7.8 【assignments for this session...

Documents