Focus Topics
An MDSI Focus Topic is a dynamically set up group of scientists at all stages (professors to doctoral researchers or even motivated master’s students) that intend to collaborate on a specific research topic along the themes of data science, machine learning or AI. Focus topics are being assembled bottom-up rather than top-down and their members define the mode of operation. This could be, e.g.
- collaborating to apply for larger structural programs with external funding agencies (like DFG clusters or CRCs),
- organizing (regular, international) workshops or conferences, or
- publishing papers.
While the members of an MDSI Focus Topic are affiliated with TUM, collaboration with external partners is strongly encouraged.
MDSI supports its Focus Topics by offering
- collaboration space at its premises,
- help and financial support for organizing workshops or conferences,
- funds for visiting researchers,
- seed funds for new projects (via annual calls),
- services related to research data management.
The following is a brief overview of the MDSI Focus Topics:
AI in Finance (AIF) combines research and application in the financial sector. It develops AI solutions for various topics, including automated trading, risk management, fraud detection, and sustainability. It also promotes digital financial education and lifelong learning to foster a close partnership between science and the financial industry.
Causal Inference (Cause) is dedicated to advancing the methodological understanding of causal inference, enabling the precise identification of cause-and-effect relationships beyond mere correlations. This is crucial for evidence-based decision-making in fields such as biomedicine, social sciences, climate research, and economics.
Computational Material Design Powered by Machine Learning (CMD-ML) explores how machine learning can accelerate computer-aided material development. The aim is to better model atomistic structure-property relationships and integrate ML models directly into simulations. The Atomistic Modeling Center (AMC) at TUM was established with this focus in mind.
InterConnect investigates graphs, networks, and connectivity structures across various disciplines, including biology, social sciences, transportation, and computer networks. Methods such as neural graph networks and topological data analysis are employed to gain a deeper understanding of information flows, social dynamics, and the spread of diseases.
Privacy-Preserving and Trustworthy Machine Learning (PPTML) deals with the development of methods for machine learning that ensure data protection and confidentiality. Since sensitive data in medicine or business is often not freely available, the topic focuses on technologies such as differential privacy, encrypted computation, and federated learning. This focus topic laid the foundation for one of the three German Konrad Zuse Schools, the Konrad Zuse School of Excellence in Reliable AI (relAI).