phd in mathematical models and methods in … · phd in mathematical models and methods in...

PhD in MATHEMATICAL MODELS AND METHODS INENGINEERING - 33rd cycle

Number of scholarship offered 16

Department DIPARTIMENTO DI MATEMATICA

Description of the PhD Programme

PhD in Mathematical Models and Methods in Engineering Research Description of the researcharea.

All Candidates interested in developing their research in one of the areas of interest of themembers of Department of Mathematics (seehttp://www.mate.polimi.it/index.php?view=ricerca)are invited to apply.They will be allowed to choose their specific research topic when theadmission procedure will be completed. It should be noted that the number of offeredscholarships may be increased before the end of the selection process. The PhD programMathematical Models and Methods in Engineering aims at training young researchers byproviding them with a strong mathematical background and with the ability to apply theirknowledge to the solution of real-world problems arising in various areas of science, technology,industry, finance, management, whenever advanced methods are required in analysis, design,planning, decision and control activities. PhD students carry their research both in thedevelopment of new mathematical methods and in the implementation and improvement ofadvanced techniques in connection with specific contexts and applications. Further informationcan be obtained by consulting the dedicated website http://www.mate.polimi.it/dottorato

Stampato il 14/04/2017 1 / 1


Research Field: DEVELOPMENT AND PROTOTYPING OF DISTRIBUTED CONTROLSYSTEMS FOR ELECTRIC NETWORKS BASED ON ADVANCED STATISTICAL MODELS

FOR THE ANALYSIS OF COMPLEX DATA.

Monthly net income of PhDscholarship (max 36 months)

€ 1200.0In case of a change of the welfare rates during the three-year period, the amount could be modified.

Context of the research activity

Motivation and objectives of the researchin this field

In the last decades, Smart Grids have revolutionized theworld of electricity generation, transport, and distributionfrom the micro-scale of domestic networks to the macro-scale of national and international networks. The typicalelectrical network has drifted from being centralized,hierarchical, analogical, and unidirectional in the flowtowards being distributed, meshed, highly sensorized, andbidirectional in the flow. In the field of Smart Grids, Italy isa world leader and a privileged venue for experimentationand research. ABB, a multinational industry of the sectorand funder of the PhD scholarship, is one of theprotagonists of this technological drift.In detail, the primaryobjective of this research is the development and study ofmathematical and statistical tools for the analysis of time-varying data generated by smart grid sensors, with aparticular attention to industrial switchboards andnetworks. In detail, we focus on complex data that can bemodelled as multivariate and/or functional data pointsembedded in suitable Hilbert spaces and continuouslygenerated by sensors locally installed along thenetwork.Among the possible data, we include bothquantities directly measured/detected over time (such ascurrents, voltages, active and reactive powers) andpossible synthesis indicators (such as their harmoniccontent, subharmonics and interharmonics, thecorrelations between them, transforms both in the timedomain and in the electrical coordinates).

1 / 4

Methods and techniques that will bedeveloped and used to carry out theresearch

The project will focus on the characterization of theprobabilistic properties of quantities/indicators as afunction of the characteristics of the network and of itsoperating regime; on the use and/or development ofstatistical methods for the identification and thecharacterization of the loads and of the users; on thedevelopment of statistical models for distributed networkmonitoring; and finally on the development of statisticalmodels for the prediction of loads, faults or abnormalbehaviors based on measured data.Particular attentionwill be given to the study of the performance limits of thedifferent triads: data, method, and prediction target. Aswell as to the robustness of the latter with respect topossible disturbances of the voltage quality (short andlong interruptions, etc).The research will be divided indetail in the following activities:1. Identification andcharacterization of data sources, objectives, and KPIs.2.Identification of test cases;3. For each data-objectivecouple, analysis of the methods and algorithms currentlyin use with particular emphasis on the most recentstatistical methods (functional data analysis and object-oriented data analysis);4. Possible adaptation of methodsto the problem under investigation and development ofnew methods: a. Theoretical study;b. Implementation in aprototypal R code;c. Analysis of test cases;d. Evaluationof KPIs.Activities 1 and 2 are planned for the first year ofPhD (in these first two activities the PhD student will havea marginal role), activity 3 is planned during the first halfof the second year of PhD and finally activity 4 during thesecond half of the second year and the entire third year.The research will be carried out at the MOX laboratory formodeling and scientific computing of the Department ofMathematics of Politecnico di Milano in close collaborationwith both the ABB research center in Dalmine (BG) andPhD students and researchers from other departments ofthe Politecnico di Milano (Department of Electronics,Information and Bio-engineering and Department ofEnergy) where other related research programs will takeplace.Finally, ABB will give to the PhD student thepossibility to spend the period abroad in Germany at theABB Corporate Research Center in Ladenburg.

2 / 4

Educational objectives

From the training point of view, the research project isaimed at the professional growth of a researcher which, atthe end of the PhD, will be provided with both rare butsought transversal skills (both theoretical and applied) inadvanced mathematical-statistical modeling for complexdata and with an ability of fruitfully interacting withresearchers from other disciplines from academia andindustry.

Job opportunities

At the end of the PhD program, the candidate will besuited for positions in advanced national and internationalresearch centers both in academia (in research groupsworking in the field of mathematical modeling, statistics,machine learning or process control) and industry (notexclusively in companies of the electricity sector but alsoin many industrial, financial, and marketing firms that findtheir current core business in big data and predictiveanalytics).

Composition of the research group1 Full Professors3 Associated Professors2 Assistant Professors6 PhD Students

Name of the research directors SIMONE VANTINI

Contacts

Simone Vantini ([email protected])

Additional support - Financial aid per PhD student per year (gross amount)

Housing - Foreign Students

1st year 2nd year 3rd year1500.0 €

per student0.0 €

per student0.0 €

per student

max number of financial aid available: 3, given in order of merit ..

Housing - Out-of-town residents(more than 80Km out of Milano) --

Additional information: educational activity, teaching assistantship, computer availability, desk availability,any other informations

The student will be given a desk and a computer at the Department of Mathematics. Within thefirst year he will be required to attend courses, workshops, study groups, and summer schoolsequivalent to 30 credits. During the three years, they will be also involved in teaching assistance

3 / 4

activities (about 100 hours). ABB will give to the PhD student the possibility to spend the periodabroad in Germany at the ABB Corporate Research Center in Ladenburg.

4 / 4


Research Field: DIVIDE-AND-CONQUER STRATEGIES: A STATISTICAL PERSPECTIVE ONTHE ANALYSIS OF MASSIVE, COMPLEX AND SPATIALLY DEPENDENT DATA





Within the framework of Object Oriented Spatial Statistics,the goal of this research project is the study of divide-and-conquer algorithms for the statistical analysis of massiveand complex data set, when spatial dependence is arelevant issue and distributed statistical inference is thecourse of action. Aim of the study is the development ofnew methods for prediction, classification and regression,together with a proper quantification of the uncertaintyassociated to the new statistics, based on the decouplingof the endogenous variability, due to the stochasticprocess generating the data and to the sampling design,from the exogenous variability introduced by the randomsub-sampling algorithms.


Starting point of the research will be the inferentialexamination of existing divide-and-conquer algorithms forthe analysis of spatially dependent object data. Focus willbe put on random domain decomposition algorithms, likethe Bagging Voronoi (Secchi et al. 2012, 2015),whereexogenous variability is introduced by the distributiongenerating the random domain partition.


The PhD track associated with this grant will becarried out in close collaboration with the DataScience programme in Bologna. In particular, themain part of the doctoral courses will be organized atthe Fondazione Golinelli, Bologna. After successful

1 / 3

completion of the PhD career at Politecnico di Milano,candidates will be offered a 1-year grant to completetheir PhD path with a fourth year at the Università diBologna, after which they can obtain a second PhDdiploma in Data Science.The present research will becarried in collaboration with the team of appliedstatisticians working in the Center for Analysis Decisionand Society of the Human Technopole infrastructure.Within this lively and stimulating academic and researchenvironment, the doctoral student will be educated tobecome a skilful data scientist exposed to theoretical andapplied statistical problems that are moving the frontier ofthe discipline.

Job opportunities

As Data Sciences are increasingly adopted by allcompanies in all industrial sectors, the demand for datascientists and analytics experts continues to outpacesupply and dominates both the US and the European jobmarket.


Name of the research directors Piercesare Secchi

Contacts

e-mail: [email protected]: +390223994592web-page:https://www.mate.polimi.it/?view=pp&id=85&lg=it




per student0.0 €

per student0.0 €

per student




A minimum endowment of ¿ 2.740,00 is granted to cover the expenses of research activities ofPhD candidates during their career.

2 / 3

There are various forms of financial aid for activities of support to the teaching practice.The PhDstudent is encouraged to take part in these activities, within the limits allowed by the regulations.Computer availability:1st year:individual use2nd year: individual use3rd year: individual useDesk availability:1st year:individual use2nd year: individual use3rd year: individual use

3 / 3


Research Field: MATHEMATICAL MODELS AND METHODS IN ENGINEERING





This Research Area covers all the topics of interest of thePhD program. All candidates interested in developing theirresearch in one of the areas of interest of the Departmentof Mathematics (seehttp://mate.polimi.it/index.php?view=ricerca) are invited toapply to the grants in this Area, which allow them tochoose their specific research topic when the admissionprocedure is completed.


The PhD program aims at training young researchers byproviding them with a strong mathematical backgroundand with ability to apply their knowledge to the solution ofreal-world problems that arise in various areas of science,technology, industry, finance, management, wheneveradvanced methods are required in analysis, design,planning, decision and control activities.


The PhD students carry their research both in thedevelopment of new mathematical methods and in theimplementation and improvement of advanced techniquesin connection with specific contexts and applications.Further information can be obtained by consulting thededicated website http://www.mate.polimi.it/dottorato

Job opportunitiesJob opportunities are in companies and researchagencies which are leaders in their respective fields.

Composition of the research group 29 Full Professors44 Associated Professors

1 / 2

22 Assistant Professors43 PhD Students

Name of the research directors Sabadini Irene Maria, Piercesare Secchi

Contacts

Program Chairman: Professor Sabadini Irene Maria e-mail: [email protected]: +39 ¿02/2399.4581 https://www.mate.polimi.it/?view=pp&id=114&lg=it#ann




per student0.0 €

per student0.0 €

per student




There are various forms of financial aid for activities of support to the teaching practice.The PhDstudent is encouraged to take part in these activities, within the limits allowed by the regulations.1individual PC per student + several shared PCAccess to one cluster with 32 processors and 384GB RAM, and to several multi-processor servers.1 individual desk per student

2 / 2


Research Field: MODELLI FISICO-MATEMATICI PER LA PROGETTAZIONE DI NUOVI META-MATERIALI ADATTATIVI





Questo progetto di ricerca è incentrato sullacaratterizzazione di regimi inesplorati d?instabilità neimateriali soffici e sull'utilizzo dei risultanti comportamentinon-convenzionali per la progettazione di nuovi meta-materiali con proprietà adattative.

Designing smart meta-material using instabilities tocontrol shape transformations, application to digitalfabrication techniques (e.g. 4d printing)


perturbation techniques, bifurcation theory, nonlinearelasticity, shape otpimization, continuation methods forPDEs


Job opportunitiesJob opportunities are in companies and researchagencies


Name of the research directors Pasquale Ciarletta

Contacts

[email protected]

1 / 2




per student0.0 €

per student0.0 €

per student




1st year: max 0,00 euros per student2nd year: max 1.370,00 euros per student3rd year: max1.370,00 euros per studentThere are various forms of financial aid for activities of support to the teaching practice.The PhDstudent is encouraged to take part in these activities, within the limits allowed by the regulations.1 individual PC per student + several shared PCAccess to one cluster with 32 processors and384 GB RAM, and to several multi-processor servers.1 individual desk per student

2 / 2


Research Field: NUMERICAL AND STATISTICAL TECHNIQUES FOR THE ANALYSIS OFFUNCTIONAL DATA





The program of this fellowship explores an emerging fieldof research at the interface between numerical analysisand statistics, focusing on the development of innovativemethods for the analysis of functional data. Recent yearshave seen an explosive growth in the recording ofcomplex and high-dimensional data often exhibiting afunctional nature, i.e., representable by suitable curves,surfaces, possibly defined over two-dimensionalmanifolds, as well as functions over complex volumetricdomains. This is essentially due to technologic progressthat has led to the development of many devices able toprovide two- and three-dimensional images and othermeasures of quantities of interest, captured in both spaceand time. High-dimensional functional data come from farranging fields: images of the internal structures of a bodyand measurements of biological signals provided bydiagnostic medical scanners, such as angiographies,tomographies and magnetic resonance; images of steadyor moving objects recorded by computer vision devices;next-generation genomic data; large speech databasesrecorded by automatic speech recognition devices; multi-spectral data from satellite remote sensing. The analysisof data characterized by a functional nature poses newand challenging problems to research and is fueling oneof the most fascinating and fast growing fields of modernstatistics. Moreover, the merging of advanced numericaland statistical methods is key to address complex datastructures. The Department of Mathematics offers a

1 / 3

unique opportunity to specialize in this area, thanks to itsstrong expertize in numerical analysis, scientificcomputing and statistics, and the synergy among theseresearch areas, with a consolidated experience ininterdisciplinary research programs.


The candidate will develop and study a class of estimatorsbased on regression with PDE regularization. Theestimation problems will be solved using finite elementmethods as well as more advanced numerical analysistechniques, such as isogeometric analysis based onNURBS and other advanced splines basis. This will bepursued via a creative and strong synergy of approachesfrom different mathematical disciplines, with a particularlyintense interplay of numerical analysis, scientificcomputing and statistics. The estimators will targetcomplex data structures, such as functional data over two-dimensional manifolds or complex volumetric domains.The applicative fields will range from life-sciences (andespecially neurosciences), to earth-sciences andengineering.


This PhD fellowship aims at providing a unique expertizeacross numerical analysis and statistics, in the domain ofdata science. Part of the research program will beconducted at EPFL École Polytechnique Fédérale deLausanne, under the supervision of Prof. Fabio Nobile,http://csqi.epfl.ch/

Job opportunities

The proposed program offers a highly qualifiedpreparation to work as data scientist and modeler inresearch and development centers (in the engineering,energy, geosciences and life sciences sectors), researchagencies, consulting companies, companies involved inthe analysis of complex data.


Name of the research directors LAURA SANGALLI

2 / 3

Contacts

[email protected], +39 02 2399 4554,http://mox.polimi.it/users/sangalli/




per student0.0 €

per student0.0 €

per student





3 / 3


Research Field: NUMERICAL METHODS FOR THE ELECTRICAL ACTIVITY OF VENTRICLESAND ATRIA WITH APPLICATION TO CASES OF CLINICAL INTEREST





Cardiac failure (CF) represents the principal cause ofclinical recovery in Italy after 65 years old. The morbidityand mortality due to CF are very high. In particular, agreat attention has been paid in the last year to theventricular dyssynchrony (VD) as a possible cause oraggravating factor of CF. Recently, a non-farmacologictherapy for VD has been established, i.e. the cardiacresynchronization therapy (CRT), obtained by stimulatingboth the ventricles with two electrodes inserted throughcatheters and stimulated by an electric implantabledevice, with the aim of restoring the physiological electricsynchrony. CRT, studied in several clinical trials, is aneffective strategy in view of improving the quality of lifeand the survival of patients with CF and VD. For thisreason, it is now included in all the official clinicalindications of the scientific societies. One of the electricpathologies leading to intra-ventricular dyssynchrony isthe left bundle branch block (LBBB), consisting in theblockage of the left branch in the heart conduction system(Purkinje network).When LBBB is treated with CRT, thetwo electrodes are usually located on the endocardium ofthe right ventricle, at the level of the intra-ventricularseptum (right electrode) and in one of the left coronaryveins (left electrode). The choice of the site of stimulationby means of the electrodes is crucial in order to obtain asyncronized contraction of the left ventricle. While the siteof stimulation of the right electrode is standard, a greatdegree of uncertainty is still present for the left one, and

1 / 4

clinicians should choice the best location for the patient athand. Indeed, recent studies have shown that thevariability of the electrical propagation in presence ofLBBB is very large, so that it is not possible to identify apriori a standard optimal choice for the location of the leftelectrode. In order to provide useful indications to theclinicians, it has been recently proposed to realize anelectrical mapping of the coronary veins (left epicardicmapping) during the sinus rhythm, with the aim ofidentifying the point characterized by the latest activation.This point is chosen for the positioning of the leftelectrode. However, this choice is nowadays still empiricaland we are far to consider it as optimal. The aim of thisproject consists in acquiring and utilizing data comingfrom the electrical mapping in patients with LBBB anddeveloping computational tools with the final goal ofoptimizing the location of the left electrode during CRT .


The units involved in this project are: APSS Unit -Divisione di Cardiologia dell'Ospedale S. Maria delCarmine di Rovereto and Dipartimento diRadiodiagnostica della APSS del Trentino; POLIMI Unit -MOX, Dipartimento di Matematica, Politecnico di Milano.APSS Unit will study patients with suspected VD,acquiring for each of them morphological data by meansof MRI or CT, ECG, and data of the electrical activation onthe endocardium and epicardium by means of the NavXsystem. These data will be given to the POLIMI Unit,which will process them allowing to obtain thecomputational geometries and the electrical maps.Starting from these elaborated data, POLIMI Unit willrealize a computational study with the aim of investigatingthe effectiveness of CRT for different left electrodelocations. To this aim, numerical approximation of partialdifferential equations to describe the electro-mechanicalproblem characterizing the ventricle function will beconsidered. In particular, the Finite Elements methods willbe used to solve such problems in the patient-specificscenarios obtained by the process of the clinical data. Inparticular, the following scenarios will be compared: 1)Healthy case. This will be virtually simulated byeliminating the block and the regions of damaged

2 / 4

propagation; 2) Sinus case. This is the "true" case, i.e. inpresence of LBBB; 3) CRT with left electrode located inthe point with the latest activation during the sinus rhythm;4) CRT with left electrode located in the point with thelatest activation during a stimulation with the rightelectrode solely; 5) Optimized CRT, i.e. with left electrodelocated in the best point (among the anatomicaladmissible ones) which guarantees the bestsynchronization. In order to find the best stimulation pointin scenario 5) and to compare its performance with thoseobtained in scenarios 3) and 4), we propose two optimalitycriteria. The first one is purely electric and it is based onthe difference between the curves representing the tissueactivation as a function of time in the healthy and CRTcases. The second one, more sophisticated, accountsalso for the mechanical behavior of the ventricle and it isbased on the capacity of systolic contraction and diastolicfilling. This will be possible thanks to a computationalstudy based on an electro-mechanical modeling.


The PhD student will learn new knowledge in differentfields: the numerical modeling of the electro-mechanicalactivityof the heart, techniques for the processing ofbiomedical data, optimization strategies, implementationin a Finite Elements library.

Job opportunities

Several opportunities will be possible: academic researchin numerical analysis and bioengineering, companies forthe development of biomedical devices or instruments,assistance to the clinicians during the operations for theuse of the biomedical diagnostic machines.


Name of the research directors CHRISTIAN VERGARA

Contacts

[email protected]

3 / 4




per student0.0 €

per student0.0 €

per student




During this PhD, the student should attend 6 courses offered by the School and teach asassistance in engineering courses. Several meeting at the hospital in Rovereto will be planned inorder to discuss with clinicians and exchange the data. 1st year: max 0,00 euros per student2ndyear: max 1.370,00 euros per student3rd year: max 1.370,00 euros per student There are variousforms of financial aid for activities of support to the teaching practice.The PhD student isencouraged to take part in these activities, within the limits allowed by the regulations. There arevarious forms of financial aid for activities of support to the teaching practice.The PhD student isencouraged to take part in these activities, within the limits allowed by the regulations. 1 individualPC per student + several shared PCAccess to one cluster with 32 processors and 384 GB RAM,and to several multi-processor servers. 1 individual desk per student

4 / 4


Research Field: NUMERICAL MODEL INTEGRATED WITH PC-MRI RADIOLOGICAL DATAFOR THE DEVELOPMENT OF THE ATHEROMASIC PLAQUE IN CAROTIDS





Interdisciplinary PhD GrantThe PhD research will be carried out in collaboration withresearch groups of the PhD programme in"BIOENGINEERING".See http://www.dottorato.polimi.it/index.php?id=242&L=1for further information.

Atherosclerosis is the leading cause of death in thewestern countries. Blood fluid-dynamics plays a crucialrole in its formation and will be studied by means ofaccurate and efficient computational models based on thecoupling between CFD and Fluid-Porous-Stucture-Interaction (FPSI) to model the atheromasic plaquegrowth. This will lead to a temporal multiscale modelwhere different temporal scales will be coupled, yeras forthe plaque development, seconds for the blood fluid-dynamics. With the aim of obtaining clinically relevantresults, patient-specific geometries based on MagneticResonance Imaging (MRI) will be acquired andreconstructed. Moreover, blood velocity data will beacquired and processed by means of the phase-contrastMRI technology. One of the principal aims of the projectwill be the integration of the PC-MRI data in the numericalmodel in order to provide suitable boundary conditionsand internal data for a possible validation. All thenumerical tolls will be developed within the FiniteElements library LIFEV (www.lifev.org).

1 / 3


This project is based on a strong synergy betweennumerical models and elaboration of biomedical images.The former ones allow one to provide quantitative resultsabout the blood fluid-dynamics and wall mechanicsinfluencing the plaque growth. The latter ones to introducein the mathematical model the patient-specific informationand the biochemical and molecular characterizations. Theimplementation of all the tools developed in the project willbe in the Finite Elements library LIFEV, specificallydeveloped for cardiovascula applications. During theproject several numerical techniques will be studied andimplemented, e.g. for the solution of the CFD/FPSIcoupling and to manage the different temporal scales. Theother important topic of the project will be thedevelopment of tools to elaborate the biomedical data, inparticular the blood velocity data based on PC-MRIacquisition. For the latter, it will be fundamental thecollaboration with the Radiology and Vascular SurgeryDivisions at Policlinico di Milano.


The PhD student will learn new knowledge indifferent fields: the numerical modeling of fluid and solidmechanics using the finite element method, use andimplementation of open source software libraries forscientific computing, techniques for the processing ofbiomedical data

Job opportunities

Several professional careers will be possible after thePh.D.: academic research in numerical analysis andbioengineering, professional opportunities in companiesfor the development of biomedical devices orinstruments, non-academic research, for examplegiving assistance for planning surgical procedures of forthe interpretation of imaging data


Name of the research directors Christian Vergara - Alberto Redaelli

Contacts

[email protected], [email protected], [email protected],2 / 3

[email protected]




per student0.0 €

per student0.0 €

per student




During this PhD, the student should attend 6 courses offered by the School and teach asassistance in engineering courses. Several meeting at the hospital in Rovereto will be planned inorder to discuss with clinicians and exchange the data. 1st year: max 0,00 euros per student2ndyear: max 1.370,00 euros per student3rd year: max 1.370,00 euros per student There are variousforms of financial aid for activities of support to the teaching practice.The PhD student isencouraged to take part in these activities, within the limits allowed by the regulations. There arevarious forms of financial aid for activities of support to the teaching practice.The PhD student isencouraged to take part in these activities, within the limits allowed by the regulations. 1 individualPC per student + several shared PCAccess to one cluster with 32 processors and 384 GB RAM,and to several multi-processor servers. 1 individual desk per student

3 / 3


Research Field: SMART CITIES¿ CHALLENGES: READING, PLANNING AND MANAGINGURBAN DYNAMICS THROUGH BIG DATA





Interdisciplinary PhD GrantThe PhD research will be carried out in collaboration withresearch groups of the PhD programme in "URBANPLANNING, DESIGN, AND POLICY".See http://www.dottorato.polimi.it/index.php?id=242&L=1for further information.

Global urbanization (Brenner and Schmidt, 2013) hasbeen creating a new condition of interdependence ofurban processes that is bringing public and private actorsto facing a great challenge: to make increasingly complexdecisions in a changing environment. With regard to thesecomplex trans-scalar interactions, Big Data offer a noveland rich source of information possibly useful to produceknowledge in innovative ways and helpful to improve therelationship between knowledge and action by giving amore detailed interpretation of the urban world whichcould be useful to increase the effectiveness of decision-making, governance, and management of cities. Throughthe development of advanced monitoring and predictiontechniques based on statistical analysis and integration ofbig data and traditional data, the research program willhelp to renew the relationship between knowledge andaction in making decisions that affect the territories of thecontemporary city. In this sense it plays selectively someof the challenges related to the idea of ¿¿"Smart City",especially one that looks at the city as a co-productionplaces of knowledge and action.

1 / 3


The complexity of the investigated phenomenon, theinformation potential of Big Data, and the construction oftools to support decision makers, make unavoidable theneed for an interdisciplinary research program aimed atdeveloping and testing innovative data scienceapproaches for the use of Big Data, driven from specificknowledge questions to improve the effectiveness andtransparency of government actions and urbanmanagement. For this reason this doctorate program willbe nurtured within the inter-departmental laboratoryUrbanscope, founded in 2016 and which includes DMAT,DAStU, DIG, DEIB and DESIGN. The MOX laboratory formodeling and scientific computing will support the studentin the development of advanced skills in the field of DataScience which will cover in detail: geostatistics, theanalysis of Hilbert data (functional, compositional andmultivariate), non-parametric inference, classification,data integration and programming. DAStU will support theresearch program offering tools for the interpretation ofthe processes of urban transformation, typical of UrbanStudies and useful to the construction of governmentscenarios for urban and territorial transformations. FinallyDIG will be in charge of the research program monitoringin a dual role: on the one hand, support the initialidentification of the decision-making needs in the differentphases (planning, monitoring, action); on the other one,test the usability of the results from the public and privatedecision-makers perspective.


From the training point of view, the research project isaimed at the professional growth of a researcher which, atthe end of the PhD, will be provided with both rare butsought transversal skills (both theoretical and applied) atthe intersection of data science, urban studies, andmanagement and with an ability of fruitfully interactingwith researchers from other disciplines from academiaand from the public sector.

Job opportunitiesAt the end of the PhD program, the candidate will besuited for positions in both advanced national andinternational research centers (in research groups working

2 / 3

in data science, urban studies, and management), in thepublic sector, and in industry as well (in industrial,financial, and marketing firms that find their current corebusiness in big data and predictive analytics for SmartCities).


Name of the research directors Simone Vantini-Valeria Fedeli

Contacts

Simone Vantini ([email protected])

Valeria Fedeli ([email protected])




per student0.0 €

per student0.0 €

per student





3 / 3


Research Field: STATISTICAL COMPUTING FOR MASSIVE DATA MANAGEMENT ANDANALYSIS





Life sciences create a huge amount of data andinformation stored in different data warehouses, withimmense potential for statistical learning. If properlyexploited, managed, and integrated, this information anddata may drive the evolution of epidemiology, clinicalstudies, and personalized medicine. The collection ofheterogeneous datasets for integrated analysis isincreasingly necessary to answer large-scale questionsconcerning, for example, citizen wellbeing and diseaseprogression. Since data sources are characterized bydifferent space and time distributions, efficient dataplatform methods are needed to integrate and query suchwidely varying data warehouses.


The project will develop statistical computational methodsaimed at handling both the multidimensional nature ofoutcomes of interest and the multiscale structure ofintegrated data.


The PhD track associated with this grant will becarried out in close collaboration with the DataScience programme in Bologna. In particular, themain part of the doctoral courses will be organized atthe Fondazione Golinelli, Bologna. After successfulcompletion of the PhD career at Politecnico di Milano,candidates will be offered a 1-year grant to completetheir PhD path with a fourth year at the Università di

1 / 2

Bologna, after which they can obtain a second PhDdiploma in Data Science.The present research will becarried in collaboration with the team of appliedstatisticians working in the Center for Analysis Decisionand Society of the Human Technopole infrastructure.Within this lively and stimulating academic and researchenvironment, the doctoral student will be educated tobecome a skilful data scientist exposed to theoretical andapplied statistical problems that are moving the frontier ofthe discipline.

Job opportunities

As Data Sciences are increasingly adopted by allcompanies in all industrial sectors, the demand for datascientists and analytics experts continues to outpacesupply and dominates both the US and the European jobmarket.


Name of the research directors Piercesare Secchi

Contacts

e-mail: [email protected]: +390223994592web-page:https://www.mate.polimi.it/?view=pp&id=85&lg=it




per student0.0 €

per student0.0 €

per student




There are various forms of financial aid for activities of support to the teaching practice.The PhDstudent is encouraged to take part in these activities, within the limits allowed by the regulations.Computer availability:1st year:individual use2nd year: individual use3rd year: individual useDesk availability:1st year:individual use2nd year: individual use3rd year: individual use

2 / 2


Research Field: THREE-DIMENSIONAL PHYSICS-BASED NUMERICAL SIMULATIONS OFEARTHQUAKE GROUND MOTION FOR ADVANCED SEISMIC RISK ASSESSMENT IN

ITALIAN URBAN AREAS





This project aims at proposing advanced approaches toseismic risk assessment in Italian urban areas, the noveltyof which consists of employing three-dimensional physics-based numerical simulations to determine earthquakeground motion instead of more standard empiricalapproaches. The project is truly interdisciplinary andrequires a joint effort of applied mathematicians, for thenumerical developments of high performance computingtools, and seismic engineers, for the evaluations ofseismic actions and of structural vulnerability for riskassessment.


This project aims at developing a fully-integratedapproach where we propose to employ physically-basedthree-dimensional wave propagation models to simulateearthquake ground motion time histories, which can thenbe used to provide the key input for fragility curves ofdifferent structural types and to quantify seismic risk. Inthis way, we aim at reducing the uncertainty related to twoof the most important components of seismic risk, i.e.,hazard and vulnerability. As a matter of fact, on one side amore accurate determination of ground motion hazard,especially for strong events, may be obtained, withrespect to standard empirical approaches based onground motion prediction equations. On the other side, theinput for fragility curves of different structural types, and

1 / 3

the corresponding predictions on possible structuralfailure, may become more reliable.The proposed fully-integrated approach will be tested considering severalItalian high seismicity urban areas. As computationallibrary, we will employ SPEED: an open source codejointly developed by proponents' research groups, whichis designed to simulate large-scale seismic events in three- dimensional complex media (including soil-structureinteraction effects). In 2013, SPEED has been selected asemerging application of high potential impact on industrialresearch and innovation within the European PRACE-2IPcall. For more details see http://speed.mox.polimi.it.Thepresent research project will be carried out in strongcollaboration with the seismic engineers' research grouplead by Prof. Paolucci (DICA, Politecnico di Milano). Thecollaboration among the research group of the proponentand Prof. Paolucci's group is long lasting, as testified byseveral cross-disciplinary publications, on-going researchprojects, and the jointly developed open-source codeSPEED (http://speed.mox.polimi.it).


The present project aim at training a high-level researcherwith a solid background in the mathematical modeling ofseismic wave propagation phenomena as well as inadvanced seismic risk assessment strategies.

Job opportunities

The candidate could apply for both academic positions inuniversities/research centers/research agencies as wellas non-academic jobs in industry and private insurancecompanies.


Name of the research directors PAOLA FRANCESCA ANTONIETTI

Contacts

e-mail: [email protected]. (+39) 02 23994601http://www1.mate.polimi.it/~antonietti


2 / 3



per student0.0 €

per student0.0 €

per student





3 / 3

phd in mathematical models and methods in … · phd in mathematical models and methods in...

Documents