Successful funding extension for deep learning project
The University of Bayreuth is participating in the DFG Research Priority Program "Theoretical Foundations of Deep Learning" for another three years.
Schematic representation of a class of neural networks being researched in the project "Nonlinear optimal feedback control with deep neural networks without the curse of dimension".
UBT/Lars Grüne
The research project "Nonlinear optimal feedback control with deep neural networks without the curse of dimension" will be funded for a further three years from October 2024 with almost €258,000. The project of the Chair of Applied Mathematics under the direction of Prof. Dr. Lars Grüne deals with the use of deep neural networks to solve large networked optimal control problems. These are problems in which an optimal strategy for the operation of a networked technical system is sought. Such systems can be found, for example, in the transportation sector in autonomous driving with many vehicles communicating with each other or in the coordination of many small systems for generating and storing renewable energy. The project uses mathematical methods to investigate which properties these problems must have so that they can be solved efficiently with deep neural networks.