GReAT: A Graph Regularized Adversarial Training Method

arXiv:2310.05336v2 Announce Type: replace
Abstract: This paper presents GReAT (Graph Regularized Adversarial Training), a novel regularization method designed to enhance the robust classification performance of deep learning models. Adversarial examples, characterized by subtle perturbations that can mislead models, pose a significant challenge in machine learning. Although adversarial training is effective in defending against such attacks, it often overlooks the underlying data structure. In response, GReAT integrates graph based regularization into the adversarial training process, leveraging the data's inherent structure to …

abstract adversarial adversarial examples adversarial training arxiv challenge classification cs.cv cs.lg deep learning examples graph machine machine learning novel paper performance regularization robust training type