After a successful first year of our GAMM student chapter, and hopefully an equally interesting semester for you, we want to introduce the first guest this year: Prof. Dr. Tobias Glasmachers. Many of you likely know him from his thrilling master modules about evolutionary algorithms and supervised machine learning, which stand out due to many practical exercises and well thought through explanations. Prof. Glasmachers will present about optimization of adaptive systems. The interplay between non-convex optimization and machine learning will be closer examined. This is certainly interesting for all engineers in the group, on the one hand, because optimization problems are ubiquitous in engineering in the form of e.g. parameter fitting or during the solution of a wide range of our beloved mathematical models. The presentation will be also interesting for the mathematicians in our group. Prof. Glasmachers is mathematician by training and will also talk about the theory of non-convex optimization. For my part I am already excited for the presentation about these interesting topics and the discussion!
Optimization and Machine Learning are closely intertwined fields. Machine learning is a rapidly growing field with excellent visibility due to many recent successes. Its underlying technology layer, optimization methods, does not enjoy the same public attention. Interestingly, in both fields, there exist nature-inspired methods: neural networks and evolutionary algorithms. In this talk, I will present current research going on in both fields within the research group Optimization of Adaptive Systems at the Institute for Neural Computation. Topics span a wide range from applied mathematical research all the way to real-world applications.