Veranstaltung

Topic of this Masterseminar are machine learning approaches and deep learning methods for learning of interpretable representations. These methods enable to reconstruct underlying principles from data, for example the reconstruction of generative factors of a dataset.
Starting from these methods for interpretable representations, we will discuss further methods for causal discovery, that enable the inference of causal dependencies in data.
Methods and algorithms covered include for example variational inference, contrastive learning, as well as statistical methods for factor analysis.
There will be a kick-off meeting at the beginning of the semester and 2-3 block seminars towards the end of the term.
Dates for both will still be determined.
The Masterseminar will be held in English language.