ICML 2020poster27 citations

Black-box Certification and Learning under Adversarial Perturbations

Hassan Ashtiani, Vinayak Pathak, Ruth Urner

Abstract

We formally study the problem of classification under adversarial perturbations from a learner’s perspective as well as a third-party who aims at certifying the robustness of a given black-box classifier. We analyze a PAC-type framework of semi-supervised learning and identify possibility and impossibility results for proper learning of VC-classes in this setting. We further introduce a new setting of black-box certification under limited query budget, and analyze this for various classes of predictors and perturbation. We also consider the viewpoint of a black-box adversary that aims at finding adversarial examples, showing that the existence of an adversary with polynomial query complexity can imply the existence of a sample efficient robust learner.

BibTeX
@InProceedings{pmlr-v119-ashtiani20a,
  title = 	 {Black-box Certification and Learning under Adversarial Perturbations},
  author =       {Ashtiani, Hassan and Pathak, Vinayak and Urner, Ruth},
  booktitle = 	 {Proceedings of the 37th International Conference on Machine Learning},
  pages = 	 {388--398},
  year = 	 {2020},
  editor = 	 {III, Hal Daumé and Singh, Aarti},
  volume = 	 {119},
  series = 	 {Proceedings of Machine Learning Research},
  month = 	 {13--18 Jul},
  publisher =    {PMLR},
  pdf = 	 {http://proceedings.mlr.press/v119/ashtiani20a/ashtiani20a.pdf},
  url = 	 {https://proceedings.mlr.press/v119/ashtiani20a.html},
  abstract = 	 {We formally study the problem of classification under adversarial perturbations from a learner’s perspective as well as a third-party who aims at certifying the robustness of a given black-box classifier. We analyze a PAC-type framework of semi-supervised learning and identify possibility and impossibility results for proper learning of VC-classes in this setting. We further introduce a new setting of black-box certification under limited query budget, and analyze this for various classes of predictors and perturbation. We also consider the viewpoint of a black-box adversary that aims at finding adversarial examples, showing that the existence of an adversary with polynomial query complexity can imply the existence of a sample efficient robust learner.}
}