arXivDaily arXiv每日学术速递 周一至周五更新
arXiv 2008.11281cs.LGcs.AIcs.DCstat.ML
收藏

加速异构数据和计算环境中的联邦学习

Accelerating Federated Learning in Heterogeneous Data and Computational Environments

Information Sciences Institute(信息科学研究所);University of Southern California(南加州大学)

Dimitris Stripelis, Jose Luis Ambite

摘要:

在某些情况下,与机器学习问题相关的数据分布在多个位置,这些位置由于监管、竞争或隐私原因而无法共享数据。例如,存在于用户手机中的数据、给定工业部门中的公司的制造数据或位于不同医院的医疗记录。此外,参与站点通常具有不同的数据分布和计算能力。联邦学习提供了一种在这些环境中学习所有可用数据的联合模型的方法。在本文中,我们介绍了一种新的分布式验证加权计划(DVW),评估联邦中的学习者对分布式验证集的性能。每个学习者保留一小部分(例如,5%)的本地训练样本作为验证dataset.We经验表明,DVW结果更好的性能相比,建立的方法,如FedAvg,在数据和计算异构环境下的同步和异步通信协议。

英文摘要:

There are situations where data relevant to a machine learning problem are distributed among multiple locations that cannot share the data due to regulatory, competitiveness, or privacy reasons. For example, data present in users' cellphones, manufacturing data of companies in a given industrial sector, or medical records located at different hospitals. Moreover, participating sites often have different data distributions and computational capabilities. Federated Learning provides an approach to learn a joint model over all the available data in these environments. In this paper, we introduce a novel distributed validation weighting scheme (DVW), which evaluates the performance of a learner in the federation against a distributed validation set. Each learner reserves a small portion (e.g., 5%) of its local training examples as a validation dataset and allows other learners models to be evaluated against it. We empirically show that DVW results in better performance compared to established methods, such as FedAvg, both under synchronous and asynchronous communication protocols in data and computationally heterogeneous environments.

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。

↑