On Regularization and Robustness of Deep Neural Networks

Alberto Bietti 1 Grégoire Mialon 1 Julien Mairal 1
1 Thoth - Apprentissage de modèles à partir de données massives
Inria Grenoble - Rhône-Alpes, LJK - Laboratoire Jean Kuntzmann
Abstract : In this work, we study the connection between regularization and robustness of deep neural networks by viewing them as elements of a reproducing kernel Hilbert space (RKHS) of functions and by regularizing them using the RKHS norm. Even though this norm cannot be computed, we consider various approximations based on upper and lower bounds. These approximations lead to new strategies for regularization, but also to existing ones such as spectral norm penalties or constraints, gradient penalties, or adversarial training. Besides, the kernel framework allows us to obtain margin-based bounds on adversarial generalization. We show that our new algorithms lead to empirical benefits for learning on small datasets and learning adversarially robust models. We also discuss implications of our regularization framework for learning implicit generative models.
Type de document :
Pré-publication, Document de travail
2018
Liste complète des métadonnées

https://hal.inria.fr/hal-01884632
Contributeur : Alberto Bietti <>
Soumis le : vendredi 30 novembre 2018 - 17:28:01
Dernière modification le : vendredi 7 décembre 2018 - 15:38:27

Fichier

main.pdf
Fichiers produits par l'(les) auteur(s)

Identifiants

  • HAL Id : hal-01884632, version 2
  • ARXIV : 1810.00363

Collections

Citation

Alberto Bietti, Grégoire Mialon, Julien Mairal. On Regularization and Robustness of Deep Neural Networks. 2018. 〈hal-01884632v2〉

Partager

Métriques

Consultations de la notice

47

Téléchargements de fichiers

83