Data Science

The certificate program is designed to provide you with a solid understanding of modern computational methods for analyzing, predicting, and visualizing data. You will learn fundamental statistical principles, broadly applicable statistical methods, modern prediction methods from machine learning, and optimization and randomization tools through a combination of theoretical concepts and practical application. The program aims to equip you with the skills to solve big data analytical problems and make informed decisions based on data.

Program:	Data Science
Graduation:	Participants receive a certificate from the Technical University of Munich after successfully passing the final exam.
Academic responsibility:	Prof. Dr. Matthias Scherer, Chair of Risk and Insurance, TUM Prof. Dr. Mathias Drton, Chair of Mathematical Statistics, TUM
Target audience:	Experts who want to expand or deepen their expertise in data science, e.g., because they currently hold or aspire to a position as a data scientist/analyst in consulting, finance, insurance, or technology.
Program fee:	3.500 Euro*
Language:	English
Discounts:	10% discount for TUM Alumni. Different conditions apply for mathematics doctoral students at TUM.
Dates:	21.-22.06.2024 + 05.-06.07.2024 + 19.-20.07.2024
Study location:	Munich near Garching & Online
Format & Timing:	Part-time, presence & digital, 6 days in 4 weeks
Admission requirements:	Participants should have a sound mathematics and computer science education or a closely related field.

Prof. Ph.D. Claudia Czado, Chair of Applied Mathematical Statistics/ MDSI, TUM	Prof. Dr. Mathias Drton, Chair of Mathematical Statistics/ MDSI, TUM
Dr. Stephan Haug, Chair of Mathematical Statistics, TUM	Prof. Dr. Blanka Horvath, Chair of Mathematical Finance/ MDSI, TUM
Prof. Dr. Oliver Junge, Chair of Numerics of Complex Systems, TUM	Prof. Dr. Felix Krahmer, Chair of Applied Numerical Analysis/ MDSI, TUM
Prof. Dr. Christina Kuttler, Chair of Mathematics in Life Sciences, TUM	Cláudio Mayrink Verdun, Chair of Applied Numerical Analysis, TUM
Prof. Dr. Matthias Scherer, Chair of Risk and Insurance, TUM	Prof. Dr. Elisabeth Ullmann, Chair of Scientific Computing and Uncertainty Quantification, TUM
Prof. Dr. Michael M. Wolf, Chair of Mathematical Physics, TUM

Module 1: Computing with Data

• An introduction to R, R Studio, and tidyverse
• Data management
• Data visualization
• Creating reports with markdown toolsR interfaces with other languages (julia, python)

Module 2: Statistical Foundations of Data Science

• Designing experiments and modeling data
• Linear regression
• Likelihood and Bayesian inference
• High-dimensional regression

Module 3: Part I: Basics of Supervised Learning

• Generative and discriminative approaches to classification and regression
• Logistic regression
• Generalized linear models
• Classification with logistic regression and discriminant analysis

Module 3: Part II: Basics of Unsupervised Learning

• Unsupervised Learning
• Clustering with k-means/k-medians, mixture models, stochastic block/ball models
• Dimension reduction with PCA/SVD
• Manifold Learning
• Autoencoders

Module 4, Part I: Predictive Approaches in ML

Kernel methods:

• support vector machines,
• Gaussian processes
• Decision trees

Ensemble methods:

• boosting and random forests

Module 4, Part II: Predictive Approaches in ML

Neural networks and deep learning:

• Training neural nets
• Approximation theory
• Network architectures

Reinforcement learning:

• Markov decision processes
• deep RL

Module 5: Optimization and Randomization for Large-Scale Data Analysis

• Non-linear optimization
• Convex optimization
• Stochastic gradient methods
• Randomization and sketching

Module 6: Case Studies & Final Exam

Presentation of case studies that exemplify applications in selected areas:

• Financial and Actuarial Math
• Examples from TUM Data Innovation Lab
• BioTech.

Assessment of your participation in the program in a pass/fail exam