Mi. 22.01.2020, Oberseminar Algebra
Thema: A Mathematicians View on Machine Learning. Referent: Prof. Dr. Markus Lange-Hegermann (Hochschule Ostwestfalen-Lippe), Raum: ENC-B 205, Zeit: 10:00 Uhr c.t.
A Mathematicians View on Machine Learning.
Mittwoch, den 22.01.2020, 10:00 Uhr c.t., Raum: ENC-B 205,
Referent: Prof. Dr. Markus Lange-Hegermann (Hochschule Ostwestfalen-Lippe)
Abstrakt: Machine Learning has exploded in the last decade, improving the state of the art in many problems or solving some problems completely. Prominent examples from the media are image recognition, natural language processing, playing games intelligently, or creating realistic data or speech. Currently, research is supported by companies and government on a large scale. In this talk, I will present a few classical ideas and recent advances in machine learning, mostly from the point of view of the underlying mathematics. In particular, I discuss the kernel trick to make linear algorithms non-linear, some ideas in deep learning, symmetries used in image recognition, how to make learning stable and automatic, and how to encode previous knowledge into learning algorithms.