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 2103.12913cs.LGstat.ML
收藏

基于半空间的$\ell_p$ -范数距离下浓度估计的改进

Improved Estimation of Concentration Under $\ell_p$-Norm Distance Metrics Using Half Spaces

University of Virginia(弗吉尼亚大学)

Jack Prescott, Xiao Zhang, David Evans

摘要:

措施的集中被认为是对抗性脆弱性的根本原因。Mahloujifar等人提出了一种使用样本来测量数据分布集中度的经验方法,并使用它来寻找几个基准数据集的内在鲁棒性下限。然而,目前还不清楚这些下限是否足够紧,以提供一个有用的近似数据集的内在鲁棒性。为了更深入地理解测度集中现象,我们首先将高斯等周不等式推广到非球面高斯测度和任意$\ell_p$ -范数($p \geq 2$)上.我们利用这些理论的见解,设计一种方法,使用半空间估计浓度下的任何经验数据集$\ell_p$ -范数距离度量。我们提出的算法是更有效的Mahloujifar等人。的,我们的实验上的合成数据集和图像基准测试表明,它是能够找到更紧密的内在鲁棒性界限。这些更严格的估计提供了进一步的证据,排除了内在数据集集中作为最先进分类器对抗性脆弱性的可能解释。

英文摘要:

Concentration of measure has been argued to be the fundamental cause of adversarial vulnerability. Mahloujifar et al. presented an empirical way to measure the concentration of a data distribution using samples, and employed it to find lower bounds on intrinsic robustness for several benchmark datasets. However, it remains unclear whether these lower bounds are tight enough to provide a useful approximation for the intrinsic robustness of a dataset. To gain a deeper understanding of the concentration of measure phenomenon, we first extend the Gaussian Isoperimetric Inequality to non-spherical Gaussian measures and arbitrary $\ell_p$-norms ($p \geq 2$). We leverage these theoretical insights to design a method that uses half-spaces to estimate the concentration of any empirical dataset under $\ell_p$-norm distance metrics. Our proposed algorithm is more efficient than Mahloujifar et al.'s, and our experiments on synthetic datasets and image benchmarks demonstrate that it is able to find much tighter intrinsic robustness bounds. These tighter estimates provide further evidence that rules out intrinsic dataset concentration as a possible explanation for the adversarial vulnerability of state-of-the-art classifiers.

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。

↑