arXivDaily arXiv每日学术速递 周一至周五更新
arXiv 2210.16258cs.CRcs.AI
收藏

多成员推断攻击下数据点的脆弱性及目标模型研究

On the Vulnerability of Data Points under Multiple Membership Inference Attacks and Target Models

University of Padua(帕多瓦大学);TU Delft(代尔夫特理工大学);University of Washington(华盛顿大学);Radboud University(拉德堡德大学)

Mauro Conti, Jiaxin Li, Stjepan Picek

摘要:

成员推断攻击(MIA)推断数据点是否在机器学习模型的训练数据中。在训练数据是一个数据点的私人信息时是一个威胁。MIA正确地推断某些数据点是训练数据的成员或非成员。直觉上,MIA准确检测到的数据点是脆弱的。考虑到这些数据点可能存在于易受多个MIA影响的不同目标模型中,多个MIA和目标模型下的数据点的脆弱性值得探索。本文定义了新的度量指标,能够反映数据点的实际脆弱性情况,捕捉多个MIA和目标模型下的脆弱数据点。分析表明,尽管整体推理性能较低,但MIA对某些数据点具有推理倾向。此外,我们实现了54个MIA,其平均攻击精度范围从0.5到0.9,以支持我们的分析与我们的可扩展和灵活的平台,成员推断攻击平台(VMIAP)。此外,以前的方法不适合在多个MIA和不同目标模型下发现脆弱数据点。最后,我们观察到漏洞不是数据点的特征,而是与MIA和目标模型相关。

英文摘要:

Membership Inference Attacks (MIAs) infer whether a data point is in the training data of a machine learning model. It is a threat while being in the training data is private information of a data point. MIA correctly infers some data points as members or non-members of the training data. Intuitively, data points that MIA accurately detects are vulnerable. Considering those data points may exist in different target models susceptible to multiple MIAs, the vulnerability of data points under multiple MIAs and target models is worth exploring. This paper defines new metrics that can reflect the actual situation of data points' vulnerability and capture vulnerable data points under multiple MIAs and target models. From the analysis, MIA has an inference tendency to some data points despite a low overall inference performance. Additionally, we implement 54 MIAs, whose average attack accuracy ranges from 0.5 to 0.9, to support our analysis with our scalable and flexible platform, Membership Inference Attacks Platform (VMIAP). Furthermore, previous methods are unsuitable for finding vulnerable data points under multiple MIAs and different target models. Finally, we observe that the vulnerability is not characteristic of the data point but related to the MIA and target model.

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。

↑