On the Existence of The Adversarial Bayes ClassifierDownload PDF

Published: 09 Nov 2021, Last Modified: 05 May 2023NeurIPS 2021 SpotlightReaders: Everyone
Keywords: Adversarial Learning, Learning Theory, Consistency, Calibration, Bayes classifiers
TL;DR: We prove that in many typical adversarial learning scenarios, there exists a classifier that minimizes the adversarial loss
Abstract: Adversarial robustness is a critical property in a variety of modern machine learning applications. While it has been the subject of several recent theoretical studies, many important questions related to adversarial robustness are still open. In this work, we study a fundamental question regarding Bayes optimality for adversarial robustness. We provide general sufficient conditions under which the existence of a Bayes optimal classifier can be guaranteed for adversarial robustness. Our results can provide a useful tool for a subsequent study of surrogate losses in adversarial robustness and their consistency properties.
Code Of Conduct: I certify that all co-authors of this work have read and commit to adhering to the NeurIPS Statement on Ethics, Fairness, Inclusivity, and Code of Conduct.
Supplementary Material: pdf
14 Replies

Loading