Machine Learning, Causal Discovery, and Experimentation (MGT001498, English) (Limited places)
| Lecturer (assistant) | |
|---|---|
| Number | MGT001498S |
| Type | seminar |
| Duration | 4 SWS |
| Term | Wintersemester 2025/26 |
| Language of instruction | English |
| Position within curricula | See TUMonline |
| Dates | See TUMonline |
Dates
-
(No dates found)
Admission information
Objectives
At the end of the seminar, the students can use Directed Acyclic Graphs (DAGs) to visual and theoretical understand causal mechanisms. They will be able to implement Generative Adversarial Networks for the creation of synthetic data and to design simple A/B experiments for business purposes. The students will learn how to assess the limitations of these approaches and how to interpret the econometric results in a meaningful way.
Description
Directed Acyclic Graphs (DAGs). Causal discovery. Conditional independence . Neural Networks. Generative Adversarial Networks. Generation of synthetic data. Causal Hypothesis Generation. Application to judges decisions in the US. Randomized Control Trials. AB experiments.
Prerequisites
First course on Statistics/Econometrics, some experience with R/Python.
Examination
Students learning outcomes will be assesed via presentations of the reading material (50% of final grade; approx. 15 presentation plus participation in the discussions during the seminar) and short written assignment (50% of final grade; approx. 10 pages without figures and tables).