Ongoing Theses

In the context of probabilistic modelling and inference as well as decision making

• Methods for determining the similarity between probabilistic graphical models (BA)
• DecPOMDPs: Implementing and evaluating a memory-bounded solution approach (MA)
• Leveraging symmetries in probabilistic circuits (BA)
• A Comparison of Different Learning Strategies for Probabilistic Circuits Based on Expressiveness and Tractability

In the context of decision making

• Recent advances in model explanations for human-aware planning (BA)

In the context of security & privacy

• Privacy-Enhanced IoT: Leveraging Oblivious Computing on Resource-Limited Devices (MA)

In the context of medical data

• In cooperation with the Institute of Neuropathology: Predicting a Gene Mutation from DNA Methylation Profiles (BA + MA)
• Modelling a Nanoscale Medical System as Decentralised Markov Decision Process (MA)

In collaboration with Fraunhofer Research Institution for Battery Cell Production FFB

• AI-based optimisation of the start-up of battery cell production