arXivDaily arXiv每日学术速递 周一至周五更新
arXiv 2102.07091math.OCcs.LGcs.SYeess.SY
收藏

Stiefel流形上的分散黎曼梯度下降

Decentralized Riemannian Gradient Descent on the Stiefel Manifold

Texas A&M University(德克萨斯农工大学);University of Minnesota(明尼苏达大学)

Shixiang Chen, Alfredo Garcia, Mingyi Hong, Shahin Shahrampour

摘要:

我们考虑一个分布式的非凸优化网络的代理人的目的是最小化一个全局函数的Stiefel流形。全局函数被表示为光滑局部函数的有限和,其中每个局部函数与一个代理相关联,并且代理通过无向连通图彼此通信。这个问题是非凸的,因为局部函数可能是非凸的(但光滑的),Steifel流形是一个非凸集。本文提出了一种分散的黎曼随机梯度法(DRSGD),其收敛速度为$\mathcal{O}(1/\sqrt{K})$。为了在恒定步长下具有精确收敛性,我们还提出了一种分散的黎曼梯度跟踪算法(DRGTA),其收敛速度为$\mathcal{O}(1/K)$到一个稳定点。我们使用多步共识来保持局部(共识)区域中的迭代。DRGTA是Stiefel流形上第一个具有精确收敛性的分布式优化算法。

英文摘要:

We consider a distributed non-convex optimization where a network of agents aims at minimizing a global function over the Stiefel manifold. The global function is represented as a finite sum of smooth local functions, where each local function is associated with one agent and agents communicate with each other over an undirected connected graph. The problem is non-convex as local functions are possibly non-convex (but smooth) and the Steifel manifold is a non-convex set. We present a decentralized Riemannian stochastic gradient method (DRSGD) with the convergence rate of $\mathcal{O}(1/\sqrt{K})$ to a stationary point. To have exact convergence with constant stepsize, we also propose a decentralized Riemannian gradient tracking algorithm (DRGTA) with the convergence rate of $\mathcal{O}(1/K)$ to a stationary point. We use multi-step consensus to preserve the iteration in the local (consensus) region. DRGTA is the first decentralized algorithm with exact convergence for distributed optimization on Stiefel manifold.

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。

↑