Mathematical foundations of artificial intelligence: International conference at the KU

Fields such as medicine and security services can benefit from new data science methods. However, this presupposes further development of areas such as machine learning, “the heart of what we call artificial intelligence”, says Prof. Dr. Götz Pfander, Holder of the Chair of Mathematics - Computational Science at the new Institute of Applied Mathematics, Machine Learning and Data Science (MIDS) at the KU. “This requires a fundamental mathematical understanding of the successes of such methods that has been lacking to date.” Hence the motivation of attendees of the International Conference on Computational Harmonic Analysis (ICCHA) that took place in Ingolstadt this year. Around 100 international guests discussed topics such as signal processing using graph algorithms, image processing or randomized algorithms.

Professor Pfander saw the ICCHA as a chance for establishing the MIDS on an international stage. “By hosting this conference, we could welcome experts from all continents, excluding Antarctica, in Ingolstadt. We received a lot of positive feedback not only regarding the contents but also for the conference location.” Prof. Dr. Holger Rauhut from RWTH Aachen University said he liked the good networking opportunities: “The presentations were very interesting – now after the long period of Zoom-only meetings, it was really nice to get together in an auditorium and chat over a coffee.”

Prof. Jared Tanner of the University of Oxford felt inspired by the interdisciplinary atmosphere of the ICCHA: “Machine learning is a very multifaceted field.” It brings together experts from mathematics, informatics and engineering. “And only when gathering in one place, they have an actual chance to learn how this field of research can be advanced and developed further as fast as possible.”

This view is also shared by Prof. Dr. Helmut Bölcskei from ETH Zurich: “Meanwhile, harmonic analysis is in the lucky position to also provide the theoretical basis in artificial intelligence. We now hope to take the understanding for such processes to the next stage and that we will be able to also conduct contributions in practice correspondingly.”

In the context of the ICCHA conference, members of the research association "Resource Aware Artificial Intelligence for Future Technologies" of the Friedrich-Alexander-University Erlangen-Nuremberg, the KU, the Technical University of Munich and the University of Bayreuth met for the first time. They coordinated further joint activities and workshops – an important step for expanding machine learning, says Prof. Pfander: “Machine learning is a rapidly developing, highly relevant field of research. Making the maximum contribution to this topic requires collaboration of several partners.” This is why the Bavarian HighTech Agenda provides funding for many professorships in this field in the context of the research association – as for example Prof. Dr. Felix Voigtlaender’s position as Chairholder of Reliable Machine Learning at the KU.