TU Delft
Year
print this page print this page     
NEDERLANDSENGLISH
Organization
2016/2017 Electrical Engineering, Mathematics and Computer Science Master Applied Mathematics
WI4231
Mathematical Data Science
ECTS: 6
Responsible Instructor
Name E-mail
Dr.ir. F.H. van der Meulen    F.H.vanderMeulen@tudelft.nl
Instructor
Name E-mail
Dr.ir. M.B. van Gijzen    M.B.vanGijzen@tudelft.nl
Dr. C. Kraaikamp    C.Kraaikamp@tudelft.nl
Dr. D. Kurowicka    D.Kurowicka@tudelft.nl
Contact Hours / Week x/x/x/x
0/0/3/0 (1 hr lab)
Education Period
3
Start Education
3
Exam Period
3
Course Language
English
Course Contents
Mathematical aspects of problems related to data science. Problems from data science are approached from three different angles: Numerical Analysis, Probability Theory and Statistics. Subjects that will be covered include the Singular Value Decomposition (with applications to e.g. tomography), counter-intuitive properties of random vectors in high dimensions, clustering, penalization and sparse data modeling.
Study Goals
The student
• Knows about how the challenge of huge amounts of data is viewed from various areas within mathematics (numerical analysis, applied probability, statistics)
• Can solve well-formulated problems in mathematical models for big data (conceptually and numerically)
• Can find, read and interpret scientific mathematics papers on the subject of big data and assess their relevance for a specific problem and apply the results
• Can, together with fellow students in the group, pose relevant questions to problem owners based on a limited description of the problem
• Can build a mathematical model for specific problems involving big data
• Can report and present results for mixed audience, consisting of ‘problem owners’ and mathematicians
Education Method
Lectures, (computer) exercise classes, working on applied problems consulting a staff member and a problem owner.
Literature and Study Materials
Course slides, scientific papers proposed by lectureres
Assessment
Biweekly home work (computer) exercises, data science project with presentation (all these in groups of 2 - 3 students) and a final (individual) oral exam.