arXivDaily arXiv每日学术速递 周一至周五更新
arXiv 2404.09536cs.DCcs.AIcs.CRcs.LG
收藏

无噪声隐私保护去中心化学习

Noiseless Privacy-Preserving Decentralized Learning

EPFL(洛桑联邦理工学院);Inria(法国国家信息与自动化研究所);PSL University(巴黎文理研究大学)

Sayan Biswas, Mathieu Even, Anne-Marie Kermarrec, Laurent Massoulie, Rafael Pires, Rishi Sharma, Martijn de Vos

AI总结:

针对去中心化学习中模型共享导致的隐私泄露问题,提出Shatter方法,通过虚拟节点分块传播模型,增强隐私并保持效用,理论证明收敛并降低攻击有效性。

摘要:

去中心化学习(DL)使得无需服务器且无需训练数据离开用户设备即可进行协作学习。然而,在DL中共享的模型仍可被用于推断训练数据。诸如差分隐私和安全聚合等传统防御措施在有效保护DL中的用户隐私方面存在不足,要么牺牲模型效用,要么牺牲效率。我们引入了Shatter,一种新颖的DL方法,其中节点创建虚拟节点(VNs)来代表它们传播其完整模型的块。这通过(i)防止攻击者从其他节点收集完整模型,以及(ii)隐藏产生给定模型块的原始节点的身份来增强隐私。我们从理论上证明了Shatter的收敛性,并提供了形式化分析,展示Shatter如何降低攻击的有效性,相比于节点之间交换完整模型。我们使用现有的DL算法、异构数据集以及针对三种标准隐私攻击来评估Shatter的收敛性和攻击韧性。我们的评估表明,当每个节点运行16个VNs时,Shatter不仅使这些隐私攻击不可行,而且与标准DL相比,对模型效用展现出积极影响。总之,Shatter在保持模型效用和效率的同时增强了DL的隐私性。

英文摘要:

Decentralized learning (DL) enables collaborative learning without a server and without training data leaving the users' devices. However, the models shared in DL can still be used to infer training data. Conventional defenses such as differential privacy and secure aggregation fall short in effectively safeguarding user privacy in DL, either sacrificing model utility or efficiency. We introduce Shatter, a novel DL approach in which nodes create virtual nodes (VNs) to disseminate chunks of their full model on their behalf. This enhances privacy by (i) preventing attackers from collecting full models from other nodes, and (ii) hiding the identity of the original node that produced a given model chunk. We theoretically prove the convergence of Shatter and provide a formal analysis demonstrating how Shatter reduces the efficacy of attacks compared to when exchanging full models between nodes. We evaluate the convergence and attack resilience of Shatter with existing DL algorithms, with heterogeneous datasets, and against three standard privacy attacks. Our evaluation shows that Shatter not only renders these privacy attacks infeasible when each node operates 16 VNs but also exhibits a positive impact on model utility compared to standard DL. In summary, Shatter enhances the privacy of DL while maintaining the utility and efficiency of the 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。

↑