Deep Learning Seminar

2019:

Julius Berner

Towards a Regularity Theory for ReLU Networks

Julius Berner

Paper Presentation: Gradient Descent Provably Optimizes Over-parameterized Neural Networks

Carlos Cancino-Chacón

Modeling Expressive Music Performance with Non-linear Basis Function Models

Monika Dörfler

Robustness and Generalization

Dennis Elbrächter

Universal sparsity of deep ReLU networks

Dennis Elbrächter

Inverse stability of the realization map for ReLU networks

Michael Feischl

Recurrent Neural Networks as Optimal Mesh Refinement Strategies

Massimo Fornasier

Robust and efficient identification of neural networks

Pavol Harar

Automatic Transformation of Empirical Data Distribution as an Experimental Pre-Processing Step for Neural Networks

Pavol Harar

Hands on introduction to Attention mechanism

Kory Johnson

Adaptive, Distribution-Free Prediction Intervals for Deep Neural Networks

Beyrem Khalfaoui

Input noise injection for improved learning with variants and applications on genomic and image data

Günther Koliander

Mutual Information in Machine Learning

Andrés Marafioti

Adversarial Generation of Time-Frequency Features with application in audio synthesis

Marco Mondelli

Analysis of a Two-Layer Neural Network via Displacement Convexity

Guido Montufar

Dynamics and Convergence of Weight Normalization for Training Neural Networks

Gregory Naitzat

Tropical algebra and topology of feedforward neural networks

Rafael Reisenhofer

Rademacher Complexities and Generalization Bounds

Martin Trapp

Sum-Product Networks

Felix Voigtlaender

Topological properties of the set of neural networks of fixed size

Felix Voigtlaender

Approximation theory of neural networks - from the concrete to the abstract

Timo Welti

Deep Artificial Neural Network Approximations for High-dimensional PDEs and Optimal Stopping Problems