Current Research in Data Science

In this seminar, we will read and discuss classic papers as well as seminal novel contributions on two of the currently most interesting topics in data science: deep learning & causality. In terms of deep learning, we will focus on the (largely unsolved) question why deep nets generalize so well. In terms of causality, we will read several classic papers to get acquainted with various causal concepts, and then discuss novel attempts to infer causal relations from non-experimental data.