English

Optimal Rates for Random Order Online Optimization

Machine Learning 2021-06-30 v1 Optimization and Control Machine Learning

Abstract

We study online convex optimization in the random order model, recently proposed by \citet{garber2020online}, where the loss functions may be chosen by an adversary, but are then presented to the online algorithm in a uniformly random order. Focusing on the scenario where the cumulative loss function is (strongly) convex, yet individual loss functions are smooth but might be non-convex, we give algorithms that achieve the optimal bounds and significantly outperform the results of \citet{garber2020online}, completely removing the dimension dependence and improving their scaling with respect to the strong convexity parameter. Our analysis relies on novel connections between algorithmic stability and generalization for sampling without-replacement analogous to those studied in the with-replacement i.i.d.~setting, as well as on a refined average stability analysis of stochastic gradient descent.

Keywords

Cite

@article{arxiv.2106.15207,
  title  = {Optimal Rates for Random Order Online Optimization},
  author = {Uri Sherman and Tomer Koren and Yishay Mansour},
  journal= {arXiv preprint arXiv:2106.15207},
  year   = {2021}
}
R2 v1 2026-06-24T03:42:23.059Z