Archetypal landscapes for deep neural networks [Applied Physical Sciences]

The predictive capabilities of deep neural networks (DNNs) continue to evolve to increasingly impressive levels. However, it is still unclear how training procedures for DNNs succeed in finding parameters that produce good results for such high-dimensional and nonconvex loss functions. In particular, we wish to understand why simple optimization schemes,...
Source: Proceedings of the National Academy of Sciences - Category: Science Authors: Tags: Physical Sciences Source Type: research