arXivDaily arXiv每日学术速递 周一至周五更新
arXiv 2012.09230math.OCcs.NAmath.NA
收藏

随机多块ADMM:基于ALM的视图,用于QP情况

Random multi-block ADMM: an ALM based view for the QP case

Stefano Cipolla, Jacek Gondzio

摘要:

在乘子的交替方向法(ADMM)中嵌入随机化过程作为对ADMM的直接多块推广不一定收敛的事实的补救,最近吸引了越来越多的兴趣。即使,在实践中,这种技术的引入可以\textit{mitigate} ADMM的多块扩展的发散行为,从理论的角度来看,它可以确保只是\textit{convergence in expectation},这可能不是一个很好的指标,其鲁棒性和效率。在这项工作中,分析强凸二次规划的情况下,我们解释块高斯-赛德尔扫描的多块ADMM的不精确增广拉格朗日方法的背景下进行。使用所提出的分析,我们能够概述一种替代技术,目前在文献中,支持更强的理论保证,是能够确保收敛的多块推广的ADMM方法。

英文摘要:

Embedding randomization procedures in the Alternating Direction Method of Multipliers (ADMM) has recently attracted an increasing amount of interest as a remedy to the fact that the direct multi-block generalization of ADMM is not necessarily convergent. Even if, in practice, the introduction of such techniques could \textit{mitigate} the diverging behaviour of the multi-block extension of ADMM, from the theoretical point of view, it can ensure just the \textit{convergence in expectation}, which may not be a good indicator of its robustness and efficiency. In this work, analysing the strongly convex quadratic programming case, we interpret the block Gauss-Seidel sweep performed by the multi-block ADMM in the context of the inexact Augmented Lagrangian Method. Using the proposed analysis, we are able to outline an alternative technique to those present in literature which, supported from stronger theoretical guarantees, is able to ensure the convergence of the multi-block generalization of the ADMM 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。

↑