Conference Agenda

2:00pm - 2:20pm

Foundations of Supervised Deep Learning for Inverse Problems

A. Auras¹, M. Burger^2,3, S. Kabri², M. Möller¹

¹University of Siegen; ²Helmholtz Imaging, Deutsches Elektronen-Synchrotron DESY; ³Universität Hamburg

2:20pm - 2:40pm

Combinatorial and implicit views on parameter optimization in neural networks

G. Montufar^1,2

¹University of California, Los Angeles, USA; ²Max Planck Institute for Mathematics in the Sciences

2:40pm - 3:00pm

Regularized, structure-preserving neural networks for the minimal entropy closure of the Boltzmann moment system

S. Schotthoefer¹, P. Laiu¹, C. Hauck¹, M. Frank²

¹Oak Ridge National Laboratory, USA; ²Karlsruhe Institute of Technology

3:00pm - 3:20pm

Adaptive Step Sizes for Preconditioned Stochastic Gradient Descent

F. Köhne¹, L. Kreis², A. Schiela¹, R. Herzog²

¹University of Bayreuth; ²Heidelberg University

3:20pm - 3:40pm

Non-vacuous PAC-Bayes bounds for Models under Adversarial Corruptions

W. Mustafa, P. Liznerski, A. Ledent, D. Wagner, P. Wang, M. Kloft

RPTU Kaiserslautern-Landau

3:40pm - 4:00pm

Convergence results for gradient flow and gradient descent systems in artificial neural network training

A. Ahmadova

University of Duisburg-Essen