Notes on Lipschitz Margin, Lipschitz Margin Training, and Lipschitz Margin p-Values for Deep Neural Network Classifiers

10/15/2019

∙

by   George Kesidis, et al.

∙

0

∙

share

We provide a local class purity theorem for Lipschitz continuous, half-rectified DNN classifiers. In addition, we discuss how to train to achieve classification margin about training samples. Finally, we describe how to compute margin p-values for test samples.