English

Smoothed Variable Sample-size Accelerated Proximal Methods for Nonsmooth Stochastic Convex Programs

Optimization and Control 2022-10-10 v4

Abstract

We consider minimizing f(x)=E[f(x,ω)]f(x) = \mathbb{E}[f(x,\omega)] when f(x,ω)f(x,\omega) is possibly nonsmooth and either strongly convex or convex in xx. (I) Strongly convex. When f(x,ω)f(x,\omega) is μ\mu-strongly convex in xx, we propose a variable sample-size accelerated proximal scheme (VS-APM) and apply it on fη(x)f_{\eta}(x), the (η\eta-)Moreau smoothed variant of E[f(x,ω)]\mathbb{E}[f(x,\omega)]; we term such a scheme as (m-VS-APM). We consider three settings. (a) Bounded domains. In this setting, VS-APM displays linear convergence in inexact gradient steps, each of which requires utilizing an inner (SSG) scheme. Specifically, mVS-APM achieves an optimal oracle complexity in SSG steps; (b) Unbounded domains. In this regime, under a weaker assumption of suitable state-dependent bounds on subgradients, an unaccelerated variant mVS-PM is linearly convergent; (c) Smooth ill-conditioned ff. When ff is LL-smooth and κ=L/μ1\kappa = L/\mu \ggg 1, we employ mVS-APM where increasingly accurate gradients xfη(x)\nabla_x f_{\eta}(x) are obtained by VS-APM. Notably, mVS-APM displays linear convergence and near-optimal complexity in inner proximal evaluations (upto a log factor) compared to VS-APM. But, unlike a direct application of VS-APM, this scheme is characterized by larger steplengths and better empirical behavior; (II) Convex. When f(x,ω)f(x,\omega) is merely convex but smoothable, by suitable choices of the smoothing, steplength, and batch-size sequences, smoothed VS-APM (or sVS-APM) produces sequences for which expected sub-optimality diminishes at the rate of O(1/k)\mathcal{O}(1/k) with an optimal oracle complexity of O(1/ϵ2)\mathcal{O}(1/\epsilon^2). Finally, sVS-APM and VS-APM produce sequences that converge almost surely to a solution of the original problem.

Keywords

Cite

@article{arxiv.1803.00718,
  title  = {Smoothed Variable Sample-size Accelerated Proximal Methods for Nonsmooth Stochastic Convex Programs},
  author = {Afrooz Jalilzadeh and Uday V. Shanbhag and Jose H. Blanchet and Peter W. Glynn},
  journal= {arXiv preprint arXiv:1803.00718},
  year   = {2022}
}

Comments

Stochastic Systems (2022)

R2 v1 2026-06-23T00:39:02.278Z