DiSCourse Seminar with Sébastien Court
10 July 2019, 11:00 (CEST)
Technikerstraße 13b, HSB 4
DiSCourse - The Digital Science Seminar Series on
Designing Mathematical Tools with Machine/Deep Learning Techniques
The aim of this interactive talk is to develop ideas for creating connections between Mathematics and the trendy topic of Machine/Deep Learning. We will first set the Mathematical context, made of optimal control theory, optimization algorithms, and differential equations modeling empirical phenomena. Next, we will present Machine Learning and Deep learning techniques, in order to address a general class of problems which consist in minimizing the distance between the model outputs and the data. Finally, we will give ideas to import the philosophy of Deep Learning, in order to design mathematical models and tools, such as partial differential equations, numerical schemes and algorithms.
Sébastien Court, University of Innsbruck, Department of Mathematics and DiSC