Research map / Optimization / Nonconvex and smooth optimization
SGDA with shuffling: faster convergence for nonconvex-PŁ minimax optimization
Read the paperCheck your idea against 108,967 papers
AI-extracted summary. Written by AI from the paper's title only (no summary yet). Numbers and claims can be wrong or missing context, so check the paper before you cite or build on it.
Measured effects reported in the paper
| What was changed | Where | Measured | Effect |
|---|---|---|---|
| random reshuffling (without-replacement) instead of with-replacement sampling in SGDA | simSGDA and altSGDA · quadratic games (P-L(Phi)-SC) | potential function V_lambda convergence | Helped |
| simultaneous vs alternating updates under RR | SGDA-RR · quadratic games (P-L(Phi)-SC) | potential function V_lambda convergence | No clear effect |
| step size ratio r=beta/alpha below kappa_2 makes simGDA impossible to converge; r of order kappa_2^2 gives tight upper bound K=O(kappa_1 r log(1/eps)) | full-batch simGDA · SC-SC and P-L(Phi)-P-L functions | iteration complexity | Mixed |
| mini-batch size b increased from 1 toward n (GDA) under RR | SGDA-RR · nonconvex-PL theory | convergence bound | No clear effect |
Closest papers by meaning
- High-probability complexity bounds for stochastic non-convex minimax optimization (NeurIPS 2024 · similarity 0.91)
- Fundamental Benefit of Alternating Updates in Minimax Optimization (ICML 2024 · similarity 0.90)
- Shuffling Gradient-Based Methods for Nonconvex-Concave Minimax Optimization (NeurIPS 2024 · similarity 0.90)
- Tighter Convergence Bounds for Shuffled SGD via Primal-Dual Perspective (NeurIPS 2024 · similarity 0.90)
- Improved Last-Iterate Convergence of Shuffling Gradient Methods for Nonsmooth Convex Optimization (ICML 2025 · similarity 0.89)
- Stochastic Extragradient with Flip-Flop Shuffling & Anchoring: Provable Improvements (NeurIPS 2024 · similarity 0.88)
Writing a paper in this area? scime draft is a free LaTeX editor with AI included, real-time co-authors and a check of your references and venue rules before you submit. It also imports OpenAI Prism projects.