Lower Difficulty and Better Robustness: A Bregman Divergence Perspective for Adversarial Training

Fuente: arXiv
Salvato in:
Dettagli Bibliografici
Autori principali: Wu, Zihui, Gao, Haichang, Zhou, Bingqian, Guo, Xiaoyan, Zhang, Shudong
Natura: Preprint
Pubblicazione: 2022
Soggetti:
Accesso online:
Tags: Aggiungi Tag
Nessun Tag, puoi essere il primo ad aggiungerne!!
_version_ 1866911748346347520
author Wu, Zihui
Gao, Haichang
Zhou, Bingqian
Guo, Xiaoyan
Zhang, Shudong
author_facet Wu, Zihui
Gao, Haichang
Zhou, Bingqian
Guo, Xiaoyan
Zhang, Shudong
contents In this paper, we investigate on improving the adversarial robustness obtained in adversarial training (AT) via reducing the difficulty of optimization. To better study this problem, we build a novel Bregman divergence perspective for AT, in which AT can be viewed as the sliding process of the training data points on the negative entropy curve. Based on this perspective, we analyze the learning objectives of two typical AT methods, i.e., PGD-AT and TRADES, and we find that the optimization process of TRADES is easier than PGD-AT for that TRADES separates PGD-AT. In addition, we discuss the function of entropy in TRADES, and we find that models with high entropy can be better robustness learners. Inspired by the above findings, we propose two methods, i.e., FAIT and MER, which can both not only reduce the difficulty of optimization under the 10-step PGD adversaries, but also provide better robustness. Our work suggests that reducing the difficulty of optimization under the 10-step PGD adversaries is a promising approach for enhancing the adversarial robustness in AT.
format Preprint
id arxiv_https___arxiv_org_abs_2208_12511
institution arXiv
publishDate 2022
record_format arxiv
spellingShingle Lower Difficulty and Better Robustness: A Bregman Divergence Perspective for Adversarial Training
Wu, Zihui
Gao, Haichang
Zhou, Bingqian
Guo, Xiaoyan
Zhang, Shudong
Machine Learning
In this paper, we investigate on improving the adversarial robustness obtained in adversarial training (AT) via reducing the difficulty of optimization. To better study this problem, we build a novel Bregman divergence perspective for AT, in which AT can be viewed as the sliding process of the training data points on the negative entropy curve. Based on this perspective, we analyze the learning objectives of two typical AT methods, i.e., PGD-AT and TRADES, and we find that the optimization process of TRADES is easier than PGD-AT for that TRADES separates PGD-AT. In addition, we discuss the function of entropy in TRADES, and we find that models with high entropy can be better robustness learners. Inspired by the above findings, we propose two methods, i.e., FAIT and MER, which can both not only reduce the difficulty of optimization under the 10-step PGD adversaries, but also provide better robustness. Our work suggests that reducing the difficulty of optimization under the 10-step PGD adversaries is a promising approach for enhancing the adversarial robustness in AT.
title Lower Difficulty and Better Robustness: A Bregman Divergence Perspective for Adversarial Training
topic Machine Learning
url https://arxiv.org/abs/2208.12511