Volker Tresp

Ludwig-Maximilians-Universität München (LMU Munich)

E-mail (email): volker.tresp at lmu dot de

Prospective students:  We are looking for students who are looking for master's thesis projects and who have a strong mathematical background!

Current Research Interests:

News | Research InterestsBiography  |  Students | Past Students | Awards and Honors | Tutorials | Software | Lecture | BooksPapers


Research Interests

Our team has a long tradition in machine learning for relational structured domains. Currently our focus is on learning with (temporal) knowledge graphs, where we also explore quantum computing solutions.   Our interest in cognitive AI was the reason that we are increasingly exploring multimodal data, such as texts, images, and videos.  Foundation models are becoming important for our work.  Our ultimate goal is to understand human level intelligence.

Tresp Lab

More research interests: 


Volker Tresp is a professor at Ludwig Maximilian University of Munich (LMU). He received his Diploma degree in physics from the University of Göttingen in 1984 and M.Sc., M.Phil. and Ph.D. degrees from Yale University in 1986 and 1989, respectively. During his Ph.D., he worked in Yale’s Image Processing and Analysis Group (IPAG). In 1990, he joined Siemens where he has been heading various research teams in machine learning. In 1997, he became Siemens Inventor of the Year for his innovations in neural networks research  and in 2018 became the first Siemens Distinguished Research Scientist. He revolutionized steel processing by pioneering a novel Bayesian neural network approach that cleverly integrated real-world data with simulated data from a prior solution.***  In 1994 he was a visiting scientist at the Massachusetts Institute of Technology in the Center for Biological and Computational Learning, working with the teams of Tomaso Poggio and Michael I. Jordan. He was co-editor of Advances in Neural Information Processing Systems 13.  In 2011, he was appointed professor in informatics at the LMU, where he teaches a course on machine learning and where he is leading a second research team. He is known for his work on Bayesian machine learning, in particular the Bayesian Committee Machine and his work on hierarchical learning with Gaussian processes. The IHRM, the SRM SUNS, and RESCAL are milestones in representation learning for multi-relational graphs. His team has been doing pioneering work on machine learning with knowledge graphs, temporal knowledge graphs, and scene graph analysis.  The work on the Tensor Brain reflects his interest in mathematical models for cognition and neuroscience. In 2020, he became a Fellow of the European Laboratory for Learning and Intelligent Systems (ELLIS). As co-director (with Kristian Kersting and Paolo Frasconi), he leads the ELLIS program "Semantic, Symbolic and Interpretable Machine Learning".

Renowned AI researcher, Geoff Hinton, aptly termed this amalgamation of prior data as "Priors without Prejudice." Subsequently, Siemens engineers adeptly tackled the challenges of concept drift and covariate shift, ensuring the model's adaptability to changing conditions and environments. As a result, the project achieved remarkable success, propelling the business unit to become a global leader in the process industry. To this day, it stands as one of the most significant achievements in machine learning for the process industry.

During the Golden Decade of the Conference on Neural Information Processing Systems (then NIPS, now, NeurIPS) from 1990-2000, I had 16 papers published in its proceedings, more than any researcher working in Europe. It was the most important machine learning conference of that decade and is leading, even today. We were shallower, but also deeper. Thanks to my co-authors: R. Hofmann, M. Roescheisen, J. Hollatz, S. Ahmad, R. Neuneier, M. Taniguchi, D. Ormoneit, H. G. Zimmermann, T. Briegel, J. Hollmen