提出新型分裂算法求解单调包含问题,可应用于离散最优传输。
Bregman Douglas-Rachford Splitting Method
- 基于Bregman距离构造交替方向分裂法,等价于对偶问题的ADMM。
- 算法在特定假设下收敛,其中一假设不适用于最优传输场景。
- 为离散最优传输提供新求解思路,适合优化与运筹领域研究者。
本文提出Bregman Douglas-Rachford分裂(BDRS)方法及其变体Bregman Peaceman-Rachford分裂方法,用于求解极大单调包含问题。我们证明,当应用于该问题的对偶形式时,BDRS等价于一种Bregman交替方向乘子法(ADMM)。Bregman ADMM的一个特例是指数乘子法的交替版本。据我们所知,本文提出的算法在文献中尚属首次出现。我们还讨论了如何利用这些算法求解离散最优传输(OT)问题。在某些假设条件下,我们证明了算法的收敛性,但指出其中一个假设并不适用于OT问题。
原文摘要 · Abstract (English)
In this paper, we propose the Bregman Douglas-Rachford splitting (BDRS) method and its variant Bregman Peaceman-Rachford splitting method for solving maximal monotone inclusion problem. We show that BDRS is equivalent to a Bregman alternating direction method of multipliers (ADMM) when applied to the dual of the problem. A special case of the Bregman ADMM is an alternating direction version of the exponential multiplier method. To the best of our knowledge, algorithms proposed in this paper are new to the literature. We also discuss how to use our algorithms to solve the discrete optimal transport (OT) problem. We prove the convergence of the algorithms under certain assumptions, though we point out that one assumption does not apply to the OT problem.
Thank you to arXiv for use of its open access interoperability. PaperDance 不是 arXiv 官方产品;中文卡片由大模型生成,请以原文为准。