Graphical Models are a combination of probability theory and graph theory. They provide a powerful tool for dealing with complex problems involving uncertainty and are playing an increasingly important role in the design and analysis of machine learning algorithms. In this seminar several aspects of graphical models will be explored, such as the automatic learning of graphical models from data, its application for decision support, data analysis and visualization, as well as a new and prominent class of graph models known as Relational Graphical Models.

Talks can be given in English or German.

Supervisor: Leandro B. Marinho, Artus Krohn-Grimberghe

Topics:

0. Intro (JPDs, Bayesian Networks, Markov Networks)
1. Structure Learning of Markov Networks
2. Structure Learning of Bayesian Networks
3. Markov Decision Processes
4. Dependency networks
5. Inference in Graphical Models
6. Latent Graphical Models
7. Hierarchical Graphical Models
8. Relational Graphical Models

Interested students can register for a topic from now via email to or .