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. 2211.01357
19
6

Quasi-Newton Steps for Efficient Online Exp-Concave Optimization

2 November 2022
Zakaria Mhammedi
Khashayar Gatmiry
ArXivPDFHTML
Abstract

The aim of this paper is to design computationally-efficient and optimal algorithms for the online and stochastic exp-concave optimization settings. Typical algorithms for these settings, such as the Online Newton Step (ONS), can guarantee a O(dln⁡T)O(d\ln T)O(dlnT) bound on their regret after TTT rounds, where ddd is the dimension of the feasible set. However, such algorithms perform so-called generalized projections whenever their iterates step outside the feasible set. Such generalized projections require Ω(d3)\Omega(d^3)Ω(d3) arithmetic operations even for simple sets such a Euclidean ball, making the total runtime of ONS of order d3Td^3 Td3T after TTT rounds, in the worst-case. In this paper, we side-step generalized projections by using a self-concordant barrier as a regularizer to compute the Newton steps. This ensures that the iterates are always within the feasible set without requiring projections. This approach still requires the computation of the inverse of the Hessian of the barrier at every step. However, using the stability properties of the Newton steps, we show that the inverse of the Hessians can be efficiently approximated via Taylor expansions for most rounds, resulting in a O(d2T+dωT)O(d^2 T +d^\omega \sqrt{T})O(d2T+dωT​) total computational complexity, where ω\omegaω is the exponent of matrix multiplication. In the stochastic setting, we show that this translates into a O(d3/ϵ)O(d^3/\epsilon)O(d3/ϵ) computational complexity for finding an ϵ\epsilonϵ-suboptimal point, answering an open question by Koren 2013. We first show these new results for the simple case where the feasible set is a Euclidean ball. Then, to move to general convex set, we use a reduction to Online Convex Optimization over the Euclidean ball. Our final algorithm can be viewed as a more efficient version of ONS.

View on arXiv
Comments on this paper