Small Gradient Norm Regret for Online Convex Optimization

Fuente: arXiv
Saved in:
Bibliographic Details
Main Authors: Gao, Wenzhi, He, Chang, Udell, Madeleine
Format: Preprint
Published: 2026
Subjects:
Online Access:
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1866917258846011392
author Gao, Wenzhi
He, Chang
Udell, Madeleine
author_facet Gao, Wenzhi
He, Chang
Udell, Madeleine
contents This paper introduces a new problem-dependent regret measure for online convex optimization with smooth losses. The notion, which we call the $G^\star$ regret, depends on the cumulative squared gradient norm evaluated at the decision in hindsight. We show that the $G^\star$ regret strictly refines the existing $L^\star$ (small loss) regret, and that it can be arbitrarily sharper when the losses have vanishing curvature around the hindsight decision. We establish upper and lower bounds on the $G^\star$ regret and extend our results to dynamic regret and bandit settings. As a byproduct, we refine the existing convergence analysis of stochastic optimization algorithms in the interpolation regime. Some experiments validate our theoretical findings.
format Preprint
id arxiv_https___arxiv_org_abs_2601_13519
institution arXiv
publishDate 2026
record_format arxiv
spellingShingle Small Gradient Norm Regret for Online Convex Optimization
Gao, Wenzhi
He, Chang
Udell, Madeleine
Machine Learning
Optimization and Control
This paper introduces a new problem-dependent regret measure for online convex optimization with smooth losses. The notion, which we call the $G^\star$ regret, depends on the cumulative squared gradient norm evaluated at the decision in hindsight. We show that the $G^\star$ regret strictly refines the existing $L^\star$ (small loss) regret, and that it can be arbitrarily sharper when the losses have vanishing curvature around the hindsight decision. We establish upper and lower bounds on the $G^\star$ regret and extend our results to dynamic regret and bandit settings. As a byproduct, we refine the existing convergence analysis of stochastic optimization algorithms in the interpolation regime. Some experiments validate our theoretical findings.
title Small Gradient Norm Regret for Online Convex Optimization
topic Machine Learning
Optimization and Control
url https://arxiv.org/abs/2601.13519