Deep Learning Theory Kickoff Meeting

Program

All talks will be held in the Leibniz-Saal (E1 05),
Max Planck Institute for Mathematics in the Sciences (Leipzig, Inselstraße 22).

 
Wednesday, March 27, 2019
08:30 - 09:00Registration / Coffee & Tee
09:00 - 09:15Welcome Address (Bernd Sturmfels)
09:15 - 09:45Guido Montúfar (Max Planck Institute for Mathematics in the Sciences)
Introduction Deep Learning Theory
09:45 - 10:00Coffee & Tee
10:00 - 11:00Nihat Ay (Max Planck Institute for Mathematics in the Sciences)
On the Natural Gradient for Deep Learning
11:00 - 12:00Tim Genewein (DeepMind London)
Neural Network Compression - model-capacity and parameter redundancy of neural networks
12:00 - 13:00Lunch Buffet at the Institute
13:00 - 14:00Ingo Steinwart (Universität Stuttgart)
A Sober Look at Neural Network Initializations
14:00 - 15:00Grégoire Montavon (Machine Learning, Technische Universität Berlin)
Explaining the Decisions of Deep Neural Networks
15:00 - 15:30Coffee & Tee
15:30 - 16:30Frederik Künstner (École Polytechnique Fédérale de Lausanne)
Limitations of the Empirical Fisher Approximation
16:30 - 18:00Poster Session & Coffee & Tee
18:00 - 18:30Discussions
19:00Dinner (together in Thueringer Hof, Burgstraße 19, 04109 Leipzig), departure from MPI main entrance at 18:30, 20 min walk
 
Thursday, March 28, 2019
09:00 - 10:00Coffee & Tee
10:00 - 11:00Razvan Pascanu (DeepMind London)
Looking at data efficiency in RL
11:00 - 12:00Maurice Weiler (Machine Learning Lab, University of Amsterdam)
Gauge Equivariant Convolutional Networks
12:00 - 13:00Lunch Buffet at the Institute
13:00 - 14:00Johannes Rauh (Max Planck Institute for Mathematics in the Sciences)
Synergy, redundancy and unique information
14:00 - 15:00Michael Arbel (Gatsby Computational Neuroscience Unit, University College London)
Kernel Distances for Deep Generative Models
15:00 - 16:00Poster Session & Coffee & Tee
16:00 - 17:00Yonatan Dukler (UCLA, Department of Mathematics)
Wasserstein of Wasserstein Loss for Learning Generative Models
17:00 - 18:00Poster Session & Coffee & Tee
19:00Dinner (individually)
 
Friday, March 29, 2019
09:00 - 10:00Coffee & Tee
10:00 - 11:00Nico Scherf (Max Planck Institute for Human Cognitive and Brain Sciences)
On Open Problems for Deep Learning in Biomedical Image Analysis
11:00 - 12:00Wuchen Li (UCLA, Department of Mathematics)
Wasserstein Information Geometry
12:00 - 13:00Lunch Buffet at the Institute
13:00 - 14:00Eliana Duarte (Max Planck Institute for Mathematics in the Sciences)
Discrete Statistical Models with Rational Maximum Likelihood Estimator
14:00 - 15:00Pradeep Banerjee (Max Planck Institute for Mathematics in the Sciences)
The Blackwell Information Bottleneck
15:00 - 15:30Coffee & Tee
15:30 - 16:30Luigi Malagò (Romanian Institute of Science and Technology - RIST, Cluj-Napoca)
On the Information Geometry of Word Embeddings
16:30 - 17:00Concluding Remarks

Date and Location

March 27 - 29, 2019
Max Planck Institute for Mathematics in the Sciences
Inselstraße 22
04103 Leipzig
Germany

Scientific Organizers

Guido Montúfar
MPI for Mathematics in the Sciences

Administrative Contact

Valeria Hünniger
MPI für Mathematik in den Naturwissenschaften

02.04.2019, 01:27