In his welcome address, Dr. Alexander Braun, Senior Vice President Digitalization and IT Systems and CIO at TUM, reported on the achievements and goals of TUM in the field of data science, machine learning and artificial intelligence.

MDSI Best Paper Awards and PhD Fellowships 2024

Together with Prof. Stephan Günnemann, Managing Director of the MDSI, Dr. Alexander Braun first awarded three young scientists for their outstanding publications in 2023. 
Norman Müller from Prof. Matthias Nießner's research group was awarded first place, Runner-up went to Kamilia Zaripova from Prof. Nassir Navab's research group and Arthur Kosmala from Prof. Stephan Günnemann's research group received an honorable mention. In addition to the prize money, all three doctoral students were presented with a certificate from the MDSI.

Two MDSI PhD Fellowships were awarded for the first time in 2024: Isabel Schorr will be supervised by Prof. Enkelejda Kasneci, Chair of Human-Centered Technologies for Learning, and Christian Grashei by Prof. Peter Schüffler, Computational Pathology research group. These new fellowships include funding for the doctoral students over three years.

Insight into current topics

In his keynote speech, Prof. David Egger gave a very good overview of the discovery and further development of functional energy materials. Especially in the prediction of material properties for the characterization and improvement of materials for sustainable energy conversion in batteries, solar cells and quantum materials, atomistic modeling methods such as machine learning with neural networks are used. Since June 2024, research has been bundled at the Atomistic Modeling Center (AMC) of TUM with Prof. Egger as one of the Directors.

Prof. Zeynep Akata, new Core Member of the MDSI, gave an insight into her research on Multimodal Large Language Models (MLLMs). Using numerous examples, she explained multimodal machine learning with the aim of generating images or correctly describing the content of an image, including target group-specific reasoning for decision-making. Another important aspect for all MLLMs is the identification and reduction of bias. Human-machine communication in apps in the healthcare and education sectors is an important area of application for such models.

In his annual report of the Munich Data Science Institute, Prof. Stephan Günnemann gave a detailed overview of the developments and events of the past year. A particular highlight was the aforementioned founding of the Atomistic Modeling Center from the focus topic “Computational Material Design Powered by ML”, as well as the other focus topics “PPTML & relAI”, “Causal Interference”, “Artificial Intelligence in Finance” (“AI in Finance”) and “InterConnect”. The industrial partnerships with SAP and Diehl, as well as outreach activities such as various hackathons, the “Women in Data Science Conference” and MDSI stands at the “TUM Sustainability Day” and “Open Day” at the Garching research campus should also be highlighted. Last year also saw further developments in research data management at the TUM Research Data Hub and the Heilbronn campus.

Following a discussion round on the further development of the programs at MDSI, a poster session was held in the Munich Data Science Institute premises. Over 50 scientists from the MDSI and the Konrad Zuse School of Excellence in Reliable AI relAI presented their research projects to interested colleagues and guests of the general assembly.

Alfred Spector's wealth of experience

To round off this exciting day, Dr. Alfred Spector, Visiting Scholar at MIT and Senior Advisor at Blackstone, former CTO and Head of Development at Two Sigma Investments, former Vice President of Research and Special Initiatives at Google and former Global Vicepresident of Services and Software Research at IBM and Global CTO of IBM's software business, gave a talk on “Beyond Models - Applying AI and Data Science Effectively”. He provided insights into the stages of his impressive career, discussed with the audience and encouraged them to think about the sensible use of artificial intelligence.achen.