Adversarially Robust Generalization Requires More Data
Ludwig Schmidt, Shibani Santurkar, Dimitris Tsipras et al. · arXiv (Cornell University) · 2018
Machine learning models are often susceptible to adversarial perturbations of their inputs. Even small perturbations can cause state-of-the-art classifiers with high "standard" accuracy to produce an incorrect prediction with high confidence. To better understand this phenomenon, we study adversarially robust learning from the viewpoint of generalization. We show that already in a simple natural data model, the sample complexity of robust learning can be significantly larger than that of "standard" learning. This gap is information theoretic and holds irrespective of the training algorithm or the model family. We complement our theoretical results with experiments on popular image classification datasets and show that a similar gap exists here as well. We postulate that the difficulty of training robust classifiers stems, at least partially, from this inherently larger sample complexity.
Cited by 476 works. doi:10.48550/arxiv.1804.11285 Open-access PDF
Authors
References
- Reading digits in natural images with unsupervised feature learning
- Explaining and Harnessing Adversarial Examples
- Statistical Decision Functions Which Minimize the Maximum Risk
- DeepFool: A Simple and Accurate Method to Fool Deep Neural Networks
- Universal Adversarial Perturbations
- Accessorize to a Crime
- Adversarial Examples for Evaluating Reading Comprehension Systems
- Feature Squeezing: Detecting Adversarial Examples in Deep Neural Networks
- Obfuscated Gradients Give a False Sense of Security: Circumventing Defenses to Adversarial Examples
- Learning in the Presence of Malicious Errors