Robustness and Generalization

Huan Xu; Shie Mannor

Robustness and Generalization

Machine Learning 2015-03-17 v1

Authors: Huan Xu , Shie Mannor

Abstract

We derive generalization bounds for learning algorithms based on their robustness: the property that if a testing sample is "similar" to a training sample, then the testing error is close to the training error. This provides a novel approach, different from the complexity or stability arguments, to study generalization of learning algorithms. We further show that a weak notion of robustness is both sufficient and necessary for generalizability, which implies that robustness is a fundamental property for learning algorithms to work.

Keywords

adversarial robustness generalization in machine learning generalization bounds

Cite

@article{arxiv.1005.2243,
  title  = {Robustness and Generalization},
  author = {Huan Xu and Shie Mannor},
  journal= {arXiv preprint arXiv:1005.2243},
  year   = {2015}
}

Related papers

View all related →