arXivDaily arXiv每日学术速递 周一至周五更新
arXiv 2311.11544cs.LGcs.CR
收藏

理解子群体对投毒攻击易感性的差异

Understanding Variation in Subpopulation Susceptibility to Poisoning Attacks

University of Virginia(弗吉尼亚大学)

Evan Rose, Fnu Suya, David Evans

AI总结:

本文研究了投毒攻击对不同子群体有效性的影响属性,发现数据集可分离性及干净模型与目标模型间的损失差异是关键因素,该结论在合成与高维基准数据集上均得到验证。

摘要:

机器学习容易受到投毒攻击,在这种攻击中,攻击者控制一小部分训练数据,并选择这些数据,目的是在训练好的模型中诱导出模型开发者非预期的行为。我们考虑一种现实场景,即具有插入有限数量数据点能力的对手试图控制模型在特定子群体上的行为。受先前关于随机标签翻转攻击对不同子群体具有不同有效性的观察启发,我们研究了可能影响最先进投毒攻击对不同子群体有效性的属性。对于一系列二维合成数据集,我们经验性地发现,对于可分离性较差的数据集,数据集可分离性在子群体脆弱性中起主导作用。然而,可分离性好的数据集表现出对单个子群体属性的更多依赖。我们进一步发现,一个关键的子群体属性可以通过干净模型与错误分类该子群体的目标模型在干净数据集上的损失差异来捕捉,如果损失差异较小,则子群体更容易被攻击。该属性也推广到高维基准数据集。对于Adult基准数据集,我们展示了可以找到与选定子群体集合的易感性相关的语义上有意义的子群体属性。本文的结果附有一个完全交互式的基于Web的子群体投毒攻击可视化,网址为https://uvasrg.github.io/visualizing-poisoning

英文摘要:

Machine learning is susceptible to poisoning attacks, in which an attacker controls a small fraction of the training data and chooses that data with the goal of inducing some behavior unintended by the model developer in the trained model. We consider a realistic setting in which the adversary with the ability to insert a limited number of data points attempts to control the model's behavior on a specific subpopulation. Inspired by previous observations on disparate effectiveness of random label-flipping attacks on different subpopulations, we investigate the properties that can impact the effectiveness of state-of-the-art poisoning attacks against different subpopulations. For a family of 2-dimensional synthetic datasets, we empirically find that dataset separability plays a dominant role in subpopulation vulnerability for less separable datasets. However, well-separated datasets exhibit more dependence on individual subpopulation properties. We further discover that a crucial subpopulation property is captured by the difference in loss on the clean dataset between the clean model and a target model that misclassifies the subpopulation, and a subpopulation is much easier to attack if the loss difference is small. This property also generalizes to high-dimensional benchmark datasets. For the Adult benchmark dataset, we show that we can find semantically-meaningful subpopulation properties that are related to the susceptibilities of a selected group of subpopulations. The results in this paper are accompanied by a fully interactive web-based visualization of subpopulation poisoning attacks found at https://uvasrg.github.io/visualizing-poisoning

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。

↑