24
0

Fully Stochastic Primal-dual Gradient Algorithm for Non-convex Optimization on Random Graphs

Abstract

Stochastic decentralized optimization algorithms often suffer from issues such as synchronization overhead and intermittent communication. This paper proposes a F\underline{\rm F}ully S\underline{\rm S}tochastic P\underline{\rm P}rimal D\underline{\rm D}ual gradient A\underline{\rm A}lgorithm (FSPDA) that suggests an asynchronous decentralized procedure with (i) sparsified non-blocking communication on random undirected graphs and (ii) local stochastic gradient updates. FSPDA allows multiple local gradient steps to accelerate convergence to stationarity while finding a consensual solution with stochastic primal-dual updates. For problems with smooth (possibly non-convex) objective function, we show that FSPDA converges to an O(σ/nT)\mathrm{\mathcal{O}( {\it \sigma /\sqrt{nT}} )}-stationary solution after T\mathrm{\it T} iterations without assuming data heterogeneity. The performance of FSPDA is on par with state-of-the-art algorithms whose convergence depend on static graph and synchronous updates. To our best knowledge, FSPDA is the first asynchronous algorithm that converges exactly under the non-convex setting. Numerical experiments are presented to show the benefits of FSPDA.

View on arXiv
Comments on this paper