1 paper
Michael Kohler, Adam Krzyzak
We present a theoretically well-founded deep learning algorithm for nonparametric regression. It uses over-parametrized deep neural networks with logistic activation function, whic…