Blogentry
Inaugural lecture of PD Dr. Matthias Rottmann
Matthias Rottmann reported in particular on methods with which an AI can be equipped for the unknown. In classification tasks, an AI works with a fixed set of categories and sorts all observations into them. Matthias Rottmann showed examples where the AI is convinced with great certainty that the classification is correct, but it is wrong after all. With his methods, Rottmann teaches the AI, among other things, to say: "I cannot assign this object to any of the categories" instead of classifying incorrectly.
Rottmann works on basic and application-oriented projects at Bergische Universität with a team of currently nine doctoral students. As deputy spokesperson of the Interdisciplinary Centre for Machine Learning and Data Analytics, he will continue to promote the activities of his faculty in the field of AI.
Last modified: 11.12.2023