Data Science in Mathematical Studies

Introduction

Photo: UHH / J.Lederer

Data science plays an essential role across almost all disciplines and industries, automating processes, detecting patterns in massive datasets, and assessing risks. In healthcare, data science aids in diagnosing diseases and personalizing treatments, while in finance, it helps detect fraud and optimize investments. Data science also powers AI-driven innovations like chatbots, recommendation systems, and self-driving technologies, reshaping the way we live and work.

The courses can be categorized into three focus areas:

Three focus areas

Probability and statistics

Course name (responsible instructor)	Lectures and lab sessions	Number of credits
High-dimensional statistics I (Johannes Lederer)	2+2 hours	8
High-dimensional statistics II (Johannes Lederer)	2+2 hours	12
Introduction to Extreme Value Theory (Holger Drees)	2+1 hours	6
Multivariate Extreme Value Statistics (Holger Drees)	2+1 hours	9
Introduction to Stein’s method (Chinmoy Bhattacharjee)	3+1 hours	8
Empirical processes (Natalie Neumeyer)	3+1 hours	8
Functional data analysis (Natalie Neumeyer)	2+1 hours	6
Bootstrap (Natalie Neumeyer)	2+1 hours	6
Asymptotic statistics (Natalie Neumeyer)	2+1 hours	6
Time series analysis (Natalie Neumeyer)	2+1 hours	9
Stochastic analysis (Vitalii Konarovskyi)	4+2 hours	12
Stochastic differential equations (with application to supervising learning) (Vitalii Konarovskyi)	3+1 hours	12

Optimization and numerical analysis

Course name (responsible instructor)	Lectures and lab sessions	Number of credits
Mathematical Machine Learning (Armin Iske)	2+1 hours	6
Inverse Problems and Learning (Martin Burger)	2+1 hours	6
Non-Smooth Optimization (Winnifried Wollner)	2+1 hours	6
Optimization Methods for Machine Learning (Winnifried Wollner)	2+1 hours	6
Introduction to Deep Learning* (Martin Spindler)	2+1 hours	6
Causal Machine Learning* (Martin Spindler)	2+1 hours	6

* Courses offered outside MIN; admission determined by the respective department

Data-science practice and applications

Course name (responsible instructor)	Lectures and lab sessions	Number of credits
Statistical Programming with Python* (Martin Spindler)	2+1 hours	6
Qualified courses from the natural sciences

* Courses offered outside MIN; admission determined by the respective department

To gain in-depth knowledge in mathematical data science, we recommend completing approximately 30 credits from the above list, with at least 9 credits from each of the first two focus areas. Note that not all courses are offered every year.

Contact

Prof. Dr. Johannes Lederer

University of Hamburg

Professor for Mathematics of data driven methods

Bundesstraße 55, Room T19

20146 Hamburg

Tel: +49 40 42838-9303

Email: johannes.lederer"AT"uni-hamburg.de

Office hours

by appointment