ResearchTrend.AI
  • Papers
  • Communities
  • Events
  • Blog
  • Pricing
Papers
Communities
Social Events
Terms and Conditions
Pricing
Parameter LabParameter LabTwitterGitHubLinkedInBlueskyYoutube

© 2025 ResearchTrend.AI, All rights reserved.

  1. Home
  2. Papers
  3. 1207.4421
43
19

Stochastic optimization and sparse statistical recovery: An optimal algorithm for high dimensions

18 July 2012
Alekh Agarwal
S. Negahban
Martin J. Wainwright
ArXivPDFHTML
Abstract

We develop and analyze stochastic optimization algorithms for problems in which the expected loss is strongly convex, and the optimum is (approximately) sparse. Previous approaches are able to exploit only one of these two structures, yielding an \order(\pdim/T)\order(\pdim/T)\order(\pdim/T) convergence rate for strongly convex objectives in \pdim\pdim\pdim dimensions, and an \order((\spindexlog⁡\pdim)/T)\order(\sqrt{(\spindex \log \pdim)/T})\order((\spindexlog\pdim)/T​) convergence rate when the optimum is \spindex\spindex\spindex-sparse. Our algorithm is based on successively solving a series of ℓ1\ell_1ℓ1​-regularized optimization problems using Nesterov's dual averaging algorithm. We establish that the error of our solution after TTT iterations is at most \order((\spindexlog⁡\pdim)/T)\order((\spindex \log\pdim)/T)\order((\spindexlog\pdim)/T), with natural extensions to approximate sparsity. Our results apply to locally Lipschitz losses including the logistic, exponential, hinge and least-squares losses. By recourse to statistical minimax results, we show that our convergence rates are optimal up to multiplicative constant factors. The effectiveness of our approach is also confirmed in numerical simulations, in which we compare to several baselines on a least-squares regression problem.

View on arXiv
Comments on this paper