News_1920x250_Detail

Gaussian processes and their interpretability

Professor Lange-Hegermann leads tutorial for graduate school DataNinja

In total, over 50 participants took part in the tutorial.

As part of the spring school of the graduate school DataNinja, a cooperation of eight research institutions in North Rhine-Westphalia, inIT board member Professor Markus Lange-Hegermann led a tutorial entitled "On Gaussian Processes and their Interpretability". The 90-minute tutorial primarily served to train the doctoral students in the graduate college, which is headed by Professor Barbara Hammer from Bielefeld University. Dr. Malte Schilling, also from Bielefeld University, is in charge of organizing the training of the PhD students as a group, including the Spring School.

"The tutorial was about teaching the modern techniques of machine learning and how to make sure that the learning algorithms remain understandable and interpretable by humans," reports Prof. Lange-Hegermann. He is Principal Investigator with the GAIA project in this graduate school. His doctoral student Andreas Besginow is doing his PhD on this project.

The tutorial took place as a hybrid event in Bielefeld. Twelve people, primarily doctoral students from the graduate school, listened in presence. "In addition, 40 people joined online. Here I am particularly pleased that the group was very international. We even had some listeners from outside Europe," Lange-Hegermann reports. The subsequent discussion and Q&A session also showed how interesting the topic was for those involved. Many questions were asked, including details of the practical implementation and the mathematical background in dealing with differential equations.

Following Professor Lange-Hegermann's presentation, the two PhD students Andreas Besginow from Lemgo and Jan Hüwel from Hagen organized a 90-minute on-site tutorial for the PhD students of the graduate school, who were given the task of reprogramming the topics of the presentation. Andreas Besginow and Jan Hüwel are familiar with the topics of the tutorial, since they are working as a tandem on their doctoral thesis in the field of the tutorial. This tandem is supervised by Prof. Lange-Hegermann and Prof. Christian Beecks from the University of Hagen.