tlooto

HE LOSS SURFACE AND EXPRESSIVITY OF DEEP CONVOLUTIONAL NEURAL NETWORKS

Quynh N. Nguyen, Matthias Hein

2018Unknown journal

Paper details

Journal
Unknown journal
Published
2018
Citations
0
References
67

tlooto Summary

This work shows that a sufficiently wide CNN has a well-behaved loss surface with almost no bad local minima, and suggests that both depth and width are equally important in deep learning.

Abstract

Abstract is not available.

Citation format

NGUYEN, Quynh N.; HEIN, Matthias. HE LOSS SURFACE AND EXPRESSIVITY OF DEEP CONVOLUTIONAL NEURAL NETWORKS. 2018.