Accelerated Decentralized Optimization with Local Updates for Smooth and Strongly Convex Objectives

[edit]

Hadrien Hendrikx, Francis Bach, Laurent Massoulie ;
Proceedings of Machine Learning Research, PMLR 89:897-906, 2019.

Abstract

In this paper, we study the problem of minimizing a sum of smooth and strongly convex functions split over the nodes of a network in a decentralized fashion. We propose the algorithm ESDACD, a decentralized accelerated algorithm that only requires local synchrony. Its rate depends on the condition number $\kappa$ of the local functions as well as the network topology and delays. Under mild assumptions on the topology of the graph, ESDACD takes a time $O((\tau_{\max} + \Delta_{\max})\sqrt{{\kappa}/{\gamma}}\ln(\epsilon^{-1}))$ to reach a precision $\epsilon$ where $\gamma$ is the spectral gap of the graph, $\tau_{\max}$ the maximum communication delay and $\Delta_{\max}$ the maximum computation time. Therefore, it matches the rate of SSDA, which is optimal when $\tau_{\max} = \Omega\left(\Delta_{\max}\right)$. Applying ESDACD to quadratic local functions leads to an accelerated randomized gossip algorithm of rate $O( \sqrt{\theta_{\rm gossip}/n})$ where $\theta_{\rm gossip}$ is the rate of the standard randomized gossip. To the best of our knowledge, it is the first asynchronous algorithm with a provably improved rate of convergence of the second moment of the error. We illustrate these results with experiments in idealized settings.

Related Material