arXivDaily arXiv每日学术速递 周一至周五更新
arXiv周末暂无论文更新,休息一下吧,周末愉快~~
arXiv 2203.13277cs.LG
收藏

对抗性风险的多样性视角

A Manifold View of Adversarial Risk

Rutgers University(罗格斯大学);Morgan Stanley(摩根士丹利);Stony Brook University(石溪大学);Queens College of CUNY(纽约市立大学皇后学院)

Wenjia Zhang, Yikai Zhang, Xiaoling Hu, Mayank Goswami, Chao Chen, Dimitris Metaxas

摘要:

机器学习模型的对抗风险已经被广泛研究。大多数以前的作品假设数据位于整个周围空间。我们建议从一个新的角度考虑流形假设。假设数据位于一个流形上,我们研究了两种新类型的对抗风险,正常的对抗风险,由于沿正常方向的扰动,和流形内的对抗风险,由于在流形内的扰动。我们证明了经典的对抗风险可以从两侧使用正常和流形上的对抗风险有界。我们还用一个令人惊讶的悲观案例表明,即使正常风险和流形内风险都为零,标准对抗风险也可以为非零。最后,我们用实证研究来支持我们的理论结果。我们的研究结果表明,通过只关注正常的对抗风险,可以提高分类器的鲁棒性。

英文摘要:

The adversarial risk of a machine learning model has been widely studied. Most previous works assume that the data lies in the whole ambient space. We propose to take a new angle and take the manifold assumption into consideration. Assuming data lies in a manifold, we investigate two new types of adversarial risk, the normal adversarial risk due to perturbation along normal direction, and the in-manifold adversarial risk due to perturbation within the manifold. We prove that the classic adversarial risk can be bounded from both sides using the normal and in-manifold adversarial risks. We also show with a surprisingly pessimistic case that the standard adversarial risk can be nonzero even when both normal and in-manifold risks are zero. We finalize the paper with empirical studies supporting our theoretical results. Our results suggest the possibility of improving the robustness of a classifier by only focusing on the normal adversarial risk.

机构、总结、中文摘要均由AI分析而来,可能存在错误,请以论文正文为准。

↑