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 2301.01142cs.LG
收藏

垂直联邦学习的互信息正则化

Mutual Information Regularization for Vertical Federated Learning

Institute for AI Industry Research, Tsinghua University(清华大学智能产业研究院)

Tianyuan Zou, Yang Liu, Ya-Qin Zhang

摘要:

垂直联邦学习(Vertical Federated Learning,VFL)被广泛应用于现实世界的应用中,以实现协作学习,同时保护数据隐私和安全。然而,以往的工作表明,没有标签的各方(被动方)在VFL可以推断的敏感标签信息所拥有的标签方(主动方)或执行后门攻击VFL。同时,主动方也可以从被动方推断敏感特征信息。所有这些都对VFL系统提出了新的隐私和安全挑战。我们提出了一种新的通用防御方法,限制私有原始数据之间的互信息,包括特征和标签,以及中间输出,以实现模型效用和隐私之间更好的权衡。我们将这种防御称为互信息正则化防御(MID)。我们从理论和实验上证明了我们的MID方法在抵御现有的攻击VFL,包括标签推理攻击,后门攻击和特征重建攻击的有效性。

英文摘要:

Vertical Federated Learning (VFL) is widely utilized in real-world applications to enable collaborative learning while protecting data privacy and safety. However, previous works show that parties without labels (passive parties) in VFL can infer the sensitive label information owned by the party with labels (active party) or execute backdoor attacks to VFL. Meanwhile, active party can also infer sensitive feature information from passive party. All these pose new privacy and security challenges to VFL systems. We propose a new general defense method which limits the mutual information between private raw data, including both features and labels, and intermediate outputs to achieve a better trade-off between model utility and privacy. We term this defense Mutual Information Regularization Defense (MID). We theoretically and experimentally testify the effectiveness of our MID method in defending existing attacks in VFL, including label inference attacks, backdoor attacks and feature reconstruction attacks.

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。

↑