A Kernel Perspective for Regularizing Deep Neural Networks
Alberto Bietti · Gregoire Mialon · Dexiong Chen · Julien Mairal

Tue Jun 11th 02:30 -- 02:35 PM @ Room 101

We propose a new point of view for regularizing deep neural networks by using the norm of a reproducing kernel Hilbert space (RKHS). Even though this norm cannot be computed, it admits upper and lower approximations leading to various practical strategies. Specifically, this perspective (i) provides a common umbrella for many existing regularization principles, including spectral norm and gradient penalties, or adversarial training, (ii) leads to new effective regularization penalties, and (iii) suggests hybrid strategies combining lower and upper bounds to get better approximations of the RKHS norm. We experimentally show this approach to be effective when learning on small datasets, or to obtain adversarially robust models.

Author Information

Alberto Bietti (Inria)
Gregoire Mialon (Inria)
Dexiong Chen (Inria)
Julien Mairal (Inria)

Related Events (a corresponding poster, oral, or spotlight)

More from the Same Authors