English

Unified analysis of SGD-type methods

Optimization and Control 2023-03-30 v1 Machine Learning

Abstract

This note focuses on a simple approach to the unified analysis of SGD-type methods from (Gorbunov et al., 2020) for strongly convex smooth optimization problems. The similarities in the analyses of different stochastic first-order methods are discussed along with the existing extensions of the framework. The limitations of the analysis and several alternative approaches are mentioned as well.

Keywords

Cite

@article{arxiv.2303.16502,
  title  = {Unified analysis of SGD-type methods},
  author = {Eduard Gorbunov},
  journal= {arXiv preprint arXiv:2303.16502},
  year   = {2023}
}

Comments

Part of the Encyclopedia of Optimization. 8 pages

R2 v1 2026-06-28T09:39:22.721Z