arXivDaily arXiv每日学术速递 周一至周五更新
arXiv 2201.12488cs.LGcs.DC
收藏

利用一类新的非线性聚集函数实现高效的分布式机器学习

Achieving Efficient Distributed Machine Learning Using a Novel Non-Linear Class of Aggregation Functions

Shanghai University of Electric Power(上海电力大学);Choate Rosemary Hall(乔特罗斯玛丽中学);Xiamen University(厦门大学);Yale University(耶鲁大学)

Haizhou Du, Ryan Yang, Yijian Chen, Qiao Xiang, Andre Wibisono, Wei Huang

摘要:

时变网络上的分布式机器学习(DML)可以成为新兴的分散式ML应用(如自动驾驶和无人机飞行)的推动者。然而,在现有的DML系统中,常用的加权算术平均模型聚合函数可能会导致高模型损失,低模型精度,以及对时变网络的收敛速度慢。为了解决这个问题,在本文中,我们提出了一种新的非线性类的模型聚合函数,以实现有效的DML随时间变化的网络。而不是采取线性聚合的相邻模型,因为大多数现有的研究,我们的机制使用非线性聚合,加权幂p均值(WPM),作为聚合函数的本地模型从邻居。随后的优化步骤使用由Bregman发散定义的镜像下降来进行,该Bregman发散保持收敛到最优。在本文中,我们分析了性质的WPM和严格证明我们的聚合机制的收敛性。此外,通过大量的实验,我们表明,当p > 1时,我们的设计显着提高了模型的收敛速度和DML下的时变网络的可扩展性相比,算术平均聚合函数,几乎没有额外的计算开销。

英文摘要:

Distributed machine learning (DML) over time-varying networks can be an enabler for emerging decentralized ML applications such as autonomous driving and drone fleeting. However, the commonly used weighted arithmetic mean model aggregation function in existing DML systems can result in high model loss, low model accuracy, and slow convergence speed over time-varying networks. To address this issue, in this paper, we propose a novel non-linear class of model aggregation functions to achieve efficient DML over time-varying networks. Instead of taking a linear aggregation of neighboring models as most existing studies do, our mechanism uses a nonlinear aggregation, a weighted power-p mean (WPM), as the aggregation function of local models from neighbors. The subsequent optimizing steps are taken using mirror descent defined by a Bregman divergence that maintains convergence to optimality. In this paper, we analyze properties of the WPM and rigorously prove convergence properties of our aggregation mechanism. Additionally, through extensive experiments, we show that when p > 1, our design significantly improves the convergence speed of the model and the scalability of DML under time-varying networks compared with arithmetic mean aggregation functions, with little additional computation overhead.

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。

↑