arXivDaily arXiv每日学术速递 周一至周五更新
arXiv 2009.02763cs.LGcs.AIstat.ML
收藏

基于垂直划分数据的混合差分私有联邦学习

Hybrid Differentially Private Federated Learning on Vertically Partitioned Data

Tencent Inc.(腾讯公司)

Chang Wang, Jian Liang, Mingkai Huang, Bing Bai, Kun Bai, Hao Li

摘要:

我们提出了HDP-VFL,第一个混合差分私有(DP)框架垂直联邦学习(VFL),以证明它是可能的,共同学习广义线性模型(GLM)从垂直分区的数据,只有一个可以忽略不计的成本,w.r.t.训练时间、准确性等,与理想化的非私有VFL相比。我们的工作建立在不同组织之间基于VFL的协作训练的最新进展之上,这些组织依赖于同态加密(HE)和安全多方计算(MPC)等协议来确保计算和训练的安全。特别是,分析了VFL的中间结果(IR)在通信过程中如何泄漏训练数据的隐私信息,并设计了一种基于DP的隐私保护算法来保证VFL参与者的数据机密性。我们从数学上证明了我们的算法不仅为VFL提供了效用保证,而且还提供了多层次的隐私,即DP w.r.t. IR和联合差分隐私(JDP)w.r.t.模型权重实验结果表明,我们的工作,在足够的隐私预算,是定量和定性类似的GLM,学习在理想化的非私人VFL设置,而不是增加成本的内存和处理时间在大多数以前的作品的基础上HE或MPC。如果这篇论文被接受,我们的代码将被发布。

英文摘要:

We present HDP-VFL, the first hybrid differentially private (DP) framework for vertical federated learning (VFL) to demonstrate that it is possible to jointly learn a generalized linear model (GLM) from vertically partitioned data with only a negligible cost, w.r.t. training time, accuracy, etc., comparing to idealized non-private VFL. Our work builds on the recent advances in VFL-based collaborative training among different organizations which rely on protocols like Homomorphic Encryption (HE) and Secure Multi-Party Computation (MPC) to secure computation and training. In particular, we analyze how VFL's intermediate result (IR) can leak private information of the training data during communication and design a DP-based privacy-preserving algorithm to ensure the data confidentiality of VFL participants. We mathematically prove that our algorithm not only provides utility guarantees for VFL, but also offers multi-level privacy, i.e. DP w.r.t. IR and joint differential privacy (JDP) w.r.t. model weights. Experimental results demonstrate that our work, under adequate privacy budgets, is quantitatively and qualitatively similar to GLMs, learned in idealized non-private VFL setting, rather than the increased cost in memory and processing time in most prior works based on HE or MPC. Our codes will be released if this paper is accepted.

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。

↑