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 2102.01936cs.LGcs.AIcs.DC
收藏

一种基于在线拉普拉斯近似的贝叶斯联邦学习框架

A Bayesian Federated Learning Framework with Online Laplace Approximation

Southern University of Science and Technology(南方科技大学);Huazhong Agricultural University(华中农业大学);Westlake University(西湖大学);OPPO Research(OPPO研究院);University of Maryland College Park(马里兰大学帕克分校);University of Chinese Academy of Sciences(中国科学院大学)

Liangxi Liu, Xi Jiang, Feng Zheng, Hong Chen, Guo-Jun Qi, Heng Huang, Ling Shao

AI总结:

本文提出一种基于在线拉普拉斯近似的贝叶斯联邦学习框架,在服务器端用多元高斯乘积构建全局后验以减少聚合误差,在客户端用全局后验先验损失约束本地训练以缓解本地遗忘,并在多个基准上取得最优结果。

摘要:

联邦学习(FL)允许多个客户端通过模型聚合和本地模型训练的循环来协作学习一个全局共享模型,而无需共享数据。现有的大多数联邦学习方法在不同客户端上分别训练本地模型,然后简单地对它们的参数取平均,以在服务器端获得一个集中式模型。然而,这些方法通常存在较大的聚合误差和严重的本地遗忘问题,在异构数据场景下这些问题尤为突出。为了解决这些问题,本文提出了一种新颖的联邦学习框架,该框架使用在线拉普拉斯近似在客户端和服务器端同时近似后验分布。在服务器端,采用多元高斯乘积机制来构建并最大化全局后验,从而大幅减少由本地模型之间较大差异引起的聚合误差。在客户端,设计了一种先验损失,利用服务器下发的全局后验概率参数来指导本地训练。通过绑定来自其他客户端的此类学习约束,我们的方法能够缓解本地遗忘。最终,我们在多个基准上取得了最先进的结果,清楚地证明了所提方法的优势。

英文摘要:

Federated learning (FL) allows multiple clients to collaboratively learn a globally shared model through cycles of model aggregation and local model training, without the need to share data. Most existing FL methods train local models separately on different clients, and then simply average their parameters to obtain a centralized model on the server side. However, these approaches generally suffer from large aggregation errors and severe local forgetting, which are particularly bad in heterogeneous data settings. To tackle these issues, in this paper, we propose a novel FL framework that uses online Laplace approximation to approximate posteriors on both the client and server side. On the server side, a multivariate Gaussian product mechanism is employed to construct and maximize a global posterior, largely reducing the aggregation errors induced by large discrepancies between local models. On the client side, a prior loss that uses the global posterior probabilistic parameters delivered from the server is designed to guide the local training. Binding such learning constraints from other clients enables our method to mitigate local forgetting. Finally, we achieve state-of-the-art results on several benchmarks, clearly demonstrating the advantages of the proposed method.

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。

↑