Document Type

Article - preprint

Department

Mathematics (CMC)

Publication Date

3-20-2014

Abstract

We obtain an improved finite-sample guarantee on the linear convergence of stochastic gradient descent for smooth and strongly convex objectives, improving from a quadratic dependence on the conditioning (L/µ) 2 (where L is a bound on the smoothness and µ on the strong convexity) to a linear dependence on L/µ. Furthermore, we show how reweighting the sampling distribution (i.e. importance sampling) is necessary in order to further improve convergence, and obtain a linear dependence in the average smoothness, dominating previous results. We also discuss importance sampling for SGD more broadly and show how it can improve convergence also in other scenarios. Our results are based on a connection we make between SGD and the randomized Kaczmarz algorithm, which allows us to transfer ideas between the separate bodies of literature studying each of the two methods. In particular, we recast the randomized Kaczmarz algorithm as an instance of SGD, and apply our results to prove its exponential convergence, but to the solution of a weighted least squares problem rather than the original least squares problem. We then present a modified Kaczmarz algorithm with partially biased sampling which does converge to the original least squares solution with the same exponential convergence rate.

Comments

Featured in Mathematical Programming Series A, vol. 155, iss. 1, 549–573, 2016.

Rights Information

© 2014 Needell, Srebro, Ward

Terms of Use & License Information

Terms of Use for work posted in Scholarship@Claremont.

Included in

Mathematics Commons

Share

COinS