Statistically Preconditioned Accelerated Gradient Method for Distributed Optimization

arXiv2002.10726MaRDI QIDQ6335509

Author name not available (Why is that?)

Publication date: 25 February 2020

Abstract: We consider the setting of distributed empirical risk minimization where multiple machines compute the gradients in parallel and a centralized server updates the model parameters. In order to reduce the number of communications required to reach a given accuracy, we propose a emph{preconditioned} accelerated gradient method where the preconditioning is done by solving a local optimization problem over a subsampled dataset at the server. The convergence rate of the method depends on the square root of the relative condition number between the global and local loss functions. We estimate the relative condition number for linear prediction models by studying emph{uniform} concentration of the Hessians over a bounded domain, which allows us to derive improved convergence rates for existing preconditioned gradient methods and our accelerated method. Experiments on real-world datasets illustrate the benefits of acceleration in the ill-conditioned regime.

Has companion code repository: https://github.com/nickolor/Yahoo-problem

This page was built for publication: Statistically Preconditioned Accelerated Gradient Method for Distributed Optimization

Report a bug (only for logged in users!)Click here to report a bug for this page (MaRDI item Q6335509)