Research map / Optimization / Nonconvex and smooth optimization
Katyusha: The First Direct Acceleration of Stochastic Gradient Methods
Zeyuan Allen-Zhu
Read the paperCheck your idea against 137,217 papers
AI-extracted. The measured effects below were read by AI from the paper's full text (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 |
|---|---|---|---|
| add Katyusha momentum (tau_2 = 1/2) to SVRG-style variance reduction (ablation) | Katyusha · LibSVM datasets (a9a, w8a, covtype, mnist, rcv1, sensit) | convergence speed | Improved |
| use Katyusha with fixed tau_2 = 1/2 instead of tuned competitors | Katyusha · LibSVM datasets | convergence speed vs SVRG, SAGA, Catalyst, APCG | Improved |
| use Catalyst instead of non-accelerated variance reduction | Catalyst · LibSVM datasets | convergence speed | Hurt |
Closest papers by meaning
- On the Convergence of Nesterov’s Accelerated Gradient Method in Stochastic Settings (ICML 2020 · similarity 0.92)
- Katyusha X: Simple Momentum Method for Stochastic Sum-of-Nonconvex Optimization (ICML 2018 · similarity 0.92)
- Gradient correlation is a key ingredient to accelerate SGD with momentum (ICLR 2025 · similarity 0.91)
- Convergence of a Stochastic Gradient Method with Momentum for Non-Smooth Non-Convex Optimization (ICML 2020 · similarity 0.91)
- Perfect Parallelization in Mini-Batch SGD with Classical Momentum Acceleration (NeurIPS 2026 · similarity 0.90)
- Non-convex Stochastic Composite Optimization with Polyak Momentum (ICML 2024 · similarity 0.89)
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.
