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 2008.01748math.OC
收藏

具有惰性和近似对偶的分散学习

Decentralized Learning with Lazy and Approximate Dual Gradients

Yanli Liu, Yuejiao Sun, Wotao Yin

摘要:

本文开发了在网络上进行分散式机器学习的算法,其中数据是分布式的,计算是本地化的,并且邻居之间的通信受到限制。这一领域最近的一系列研究集中在提高计算和通信复杂性上。SSDA和MSDA方法\cite{scaman2017optimal}在目标光滑和强凸时具有最优的通信复杂度,并且推导简单。然而,它们需要在每一步解决一个子问题。我们提出了新的算法,节省计算通过使用(随机)梯度和保存通信时,以前的信息是足够有用的。我们的方法仍然相对简单-而不是解决一个子问题,他们从最新的点运行卡秋莎一个小的,固定的步骤。一个易于计算的本地规则用于决定工人是否可以跳过一轮通信。此外,我们的方法可证明降低通信和计算的复杂性SSDA和MSDA。在数值实验中,我们的算法实现了显着的计算和通信减少相比,国家的最先进的。

英文摘要:

This paper develops algorithms for decentralized machine learning over a network, where data are distributed, computation is localized, and communication is restricted between neighbors. A line of recent research in this area focuses on improving both computation and communication complexities. The methods SSDA and MSDA \cite{scaman2017optimal} have optimal communication complexity when the objective is smooth and strongly convex, and are simple to derive. However, they require solving a subproblem at each step. We propose new algorithms that save computation through using (stochastic) gradients and saves communications when previous information is sufficiently useful. Our methods remain relatively simple -- rather than solving a subproblem, they run Katyusha for a small, fixed number of steps from the latest point. An easy-to-compute, local rule is used to decide if a worker can skip a round of communication. Furthermore, our methods provably reduce communication and computation complexities of SSDA and MSDA. In numerical experiments, our algorithms achieve significant computation and communication reduction compared with the state-of-the-art.

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。

↑