arXivDaily arXiv每日学术速递 周一至周五更新
arXiv 2209.08615cs.LGcs.AIcs.CR
收藏

成员推断攻击与泛化:因果视角

Membership Inference Attacks and Generalization: A Causal Perspective

National University of Singapore(新加坡国立大学);Hong Kong University of Science and Technology(香港科技大学);Microsoft Research(微软研究院)

Teodora Baluta, Shiqi Shen, S. Hitarth, Shruti Tople, Prateek Saxena

摘要:

隶属度推断(MI)攻击突出了现有神经网络随机训练方法的隐私弱点。然而,人们并不很清楚它们为什么会出现。它们仅仅是不完美的概括的自然结果吗?在培训期间,我们应该解决哪些根本原因,以减轻这些攻击?为了回答这些问题,我们提出了第一种方法来解释MI攻击及其与基于原则因果推理的泛化的联系。我们提供因果关系图,定量解释所观察到的MI攻击性能达到$6$攻击变种。我们驳斥了几个先前的非定量假设,这些假设过度简化或高估了潜在原因的影响,从而未能捕捉到几个因素之间复杂的相互作用。我们的因果模型还显示了泛化和MI攻击之间的新联系,通过其共享的因果因素。我们的因果模型具有高预测能力($0.90$),即,他们的分析预测往往与看不见的实验中的观察相吻合,这使得通过他们进行分析成为一种实用的选择。

英文摘要:

Membership inference (MI) attacks highlight a privacy weakness in present stochastic training methods for neural networks. It is not well understood, however, why they arise. Are they a natural consequence of imperfect generalization only? Which underlying causes should we address during training to mitigate these attacks? Towards answering such questions, we propose the first approach to explain MI attacks and their connection to generalization based on principled causal reasoning. We offer causal graphs that quantitatively explain the observed MI attack performance achieved for $6$ attack variants. We refute several prior non-quantitative hypotheses that over-simplify or over-estimate the influence of underlying causes, thereby failing to capture the complex interplay between several factors. Our causal models also show a new connection between generalization and MI attacks via their shared causal factors. Our causal models have high predictive power ($0.90$), i.e., their analytical predictions match with observations in unseen experiments often, which makes analysis via them a pragmatic alternative.

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。

↑