Research map / Optimization / Nonconvex and smooth optimization
A simpler approach to accelerated optimization: iterative averaging meets optimism
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 |
|---|---|---|---|
| Anytime online-to-batch iterate averaging (querying gradient at online average) combined with optimistic AO-FTRL dual averaging, tightened analysis keeping negative Bregman terms | Stochastic proximal dual averaging · Smooth non-strongly-convex stochastic optimization | optimization gap E[ℓ(x̄_T)]−ℓ⋆ rate | Helped |
| Negative momentum toward snapshot point added to variance-reduced anytime averaging (ablation) | Variance-reduced accelerated method (Algorithm 3) · Finite-sum smooth non-strongly-convex optimization | convergence rate (last iterate) | Helped |
| Removing negative momentum (no-momentum variant with snapshot = last iterate) (ablation) | Variance-reduced anytime online-to-batch · Finite-sum smooth non-strongly-convex optimization | convergence rate | Hurt |
| Adding Cutkosky-style iterate averaging and optimism together with AdaGrad-style coordinatewise step sizes (universal algorithm) | Universal composite non-strongly-convex algorithm · Smooth and non-smooth composite objectives | convergence rate without knowing smoothness/noise | Helped |
Closest papers by meaning
- Anytime Online-to-Batch, Optimism and Acceleration (ICML 2019 · similarity 0.92)
- Optimistic Online-to-Batch Conversions for Accelerated Convergence and Universality (NeurIPS 2025 · similarity 0.92)
- Optimal Stochastic Non-smooth Non-convex Optimization through Online-to-Non-convex Conversion (ICML 2023 · similarity 0.91)
- Gradient-Variation Online Adaptivity for Accelerated Optimization with Hölder Smoothness (NeurIPS 2025 · similarity 0.90)
- Accelerated Quasi-Newton Proximal Extragradient: Faster Rate for Smooth Convex Optimization (NeurIPS 2023 · similarity 0.90)
- Universality of AdaGrad Stepsizes for Stochastic Optimization: Inexact Oracle, Acceleration and Variance Reduction (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.