Whether you’re a researcher, student, or industry professional, this workshop is a great opportunity to learn about the latest developments in AI and mathematics, and connect with peers and experts in the field.
Over the course of two days, thematic sessions will highlight connections between the different clusters and provide opportunities for networking and collaboration.
We hope to see many of you in Groningen!
Registration
More information will follow.
Travel Directions
More information will follow.
Keynote Speakers
chair: Tim van Erven
Four Thematic Sessions
1. Graph Signal Processing and AI, session chair: Palina Salanevich
Graph signal processing (GSP) is a new and thriving area in mathematical data science that aims to develop analysis tools and reconstruction techniques for data supported on graphs, including social and economic networks, brain imaging, epidemiology, and traffic networks. In this session, we will discuss how the tools developed in GSP can be applied to process, analyze, and visualize large-scale structured data in machine learning.
2. Sequential Decision Making, session chair: Wouter M. Koolen
Sequential decision making is a thread connecting many areas of learning theory, including online optimization, active statistical testing, bandits, reinforcement learning, solving of games.
3. Inverse Problems and Dynamical Systems, session chair: Kerstin Bunte
Inverse problems and dynamical systems focus on reconstructing and understanding evolving processes from limited or noisy data. Inverse problems infer hidden parameters from observations, while dynamical systems describe how those parameters govern time-dependent behaviour. This session explores how combining these fields with modern AI techniques enables more accurate prediction, robust control, and deeper analysis of complex real-world phenomena.
4. Scientific Machine Learning, session chair: Dimitris Loukrezis
Scientific Machine Learning (SciML) is a relatively young field that merges traditional scientific computation with machine learning (ML). Its goal is to develop novel methods that combine physics-based knowledge in the form of, e.g., mathematical equations and physical properties, with data-driven, ML-based modeling. In this section, we will discuss methodological developments and applications of SciML in mathematics, physics, and engineering.
Program
More information will follow.
Visual impression (based on previous editions)
