The gap between theory and practice in function approximation with deep neural networks

Open in new window