This page is a work in progress. Please check it regularly for more content on our thesis topics.
Our availability to supervise Bachelor's and Master's theses in Summer Semester 2024 is closed. New opportunities for Winter Semester 24-25 and Summer Semester 2025 will be announced soon.
Theory and Methods
Robust, Data-Efficient, and Discrete-Continuous Learning
- Combining discrete probability and Deep Learning
- Explainable AI
- ML and combinatorial optimization
- ML as surrogate models
(Geometric) Deep Learning
- Graph Neural networks
- ML for multi-modal data
- ML for Spatio-temporal data, simulations, PDEs
- Equivariant NNS
Applications
- ML for Molecules
- ML for Physical Systems
- ML for (Bio-) Medical Applications
Ongoing Research at MLS
- Graph Neural Networks (GNNs) for Protein-Ligand binding prediction
Vinh Tong, M.Sc. - Towards robust and trustworthy Geometric Deep Learning
Andrei Manolache, M.Sc. - Machine Learning for Partial Differential Equations
Supervisor: Marimuthu Kalimuthu, M.Sc. - Vision and Language: Fusion Models, Flamingo
Marimuthu Kalimuthu, M.Sc. - Generative Modeling (VLM, LLM, Diffusion Models, etc.)
Marimuthu Kalimuthu, M.Sc. - Modeling Long Range Dependencies for Sequential Data
Marimuthu Kalimuthu, M.Sc.