arXivDaily arXiv每日学术速递 周一至周五更新
arXiv 2103.13822cs.LGcs.DC
收藏

FedCor:基于相关性的异构联邦学习主动客户端选择策略

FedCor: Correlation-Based Active Client Selection Strategy for Heterogeneous Federated Learning

Duke University(杜克大学);Tsinghua University(清华大学)

Minxue Tang, Xuefei Ning, Yitu Wang, Jingwei Sun, Yu Wang, Hai Li, Yiran Chen

摘要:

在联邦学习(FL)中,客户端数据异构性是阻碍有效训练的主要问题之一。由于每个客户端上的数据分布可能会有很大的不同,客户端选择策略可以显着影响FL过程的收敛速度。主动客户选择策略是近年来研究的热点。然而,他们忽略了客户端之间的损失相关性,并实现了边际改善相比,统一的选择策略。在这项工作中,我们提出了FedCor --一个建立在基于相关性的客户选择策略上的FL框架,以提高FL的收敛速度。具体来说,我们首先使用高斯过程(GP)对客户之间的损失相关性进行建模。基于GP模型,我们得到了一个客户端选择策略,在每一轮的预期全球损失显着减少。此外,我们开发了一个有效的GP训练方法,在FL场景下,通过利用协方差平稳性的低通信开销。实验结果表明,与现有方法相比,FedCorr算法在FMNIST和CIFAR-10上的收敛速度分别提高了$34\%\sim 99\%$和$26\%\sim 51\%$.

英文摘要:

Client-wise data heterogeneity is one of the major issues that hinder effective training in federated learning (FL). Since the data distribution on each client may vary dramatically, the client selection strategy can significantly influence the convergence rate of the FL process. Active client selection strategies are popularly proposed in recent studies. However, they neglect the loss correlations between the clients and achieve only marginal improvement compared to the uniform selection strategy. In this work, we propose FedCor -- an FL framework built on a correlation-based client selection strategy, to boost the convergence rate of FL. Specifically, we first model the loss correlations between the clients with a Gaussian Process (GP). Based on the GP model, we derive a client selection strategy with a significant reduction of expected global loss in each round. Besides, we develop an efficient GP training method with a low communication overhead in the FL scenario by utilizing the covariance stationarity. Our experimental results show that compared to the state-of-the-art method, FedCorr can improve the convergence rates by $34\%\sim 99\%$ and $26\%\sim 51\%$ on FMNIST and CIFAR-10, respectively.

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。

↑