On February 7, the machine learning lecture at the Faculty of Physics crossed the finish line with the last event. Instead of a written exam, the participants were involved in science-related AI projects (so-called capstone projects) which they presented in a poster session. This year, 6 projects were submitted by the HZDR. Participating institutes were FWD, FWK and FWC. The lecture itself was held by the Helmholtz AI Consultant Unit (FWCC-A). Peter Steinbach, Sebastian Starke and Steve Schmerler introduced the students to the basics of modern AI for the analysis of scientific data in the natural sciences.
Among the participating teams in the capstone projects, two were recognized with awards! Team Tom Haustein and Victor Cozea dealt with “Learning Reduced Representations for Particles in Kinetic Plasma Simulations” and won the jury prize (picture below left). Marie Neumaier and Valentin Rothert's team (picture below right) won the audience poster prize for “Fine-tuning variational autoencoders”. Both projects were provided by the Helmholtz AI Young Investigator Group around Jeffrey Kelling (FWK) and mentored by Vedhas Pandit (FWK) from the group. Thanks again to all mentors and participants of the Capstone Poster Session!
Contact: Dr. Steinbach, Peter
