Research map / Optimization / Nonconvex and smooth optimization
STORM+: Fully Adaptive SGD with Recursive Momentum for Nonconvex 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 |
|---|---|---|---|
| Use jointly adaptive learning-rate and momentum choices in place of STORM's parameter-dependent choices | STORM+ · Expectation over smooth losses, stochastic | expected gradient norm: O((log T)^{3/4}/√T + σ^{1/3}/T^{1/3}) -> O(M/√T + κσ^{1/3}/T^{1/3}) | Helped |
| Replace adaptive momentum with fixed a₁ = 1 and aₜ₊₁ = 1/t^{2/3} (ablation) | STORM+ · Expectation over smooth losses, stochastic | expected gradient norm: O(√(L³ + σ² + B^{3/2})/T^{1/3}) | Mixed |
| Run with a deterministic oracle (σ = 0) (ablation) | STORM+ · Deterministic oracle | expected gradient norm: O(√(1+L³+B^{9/4})/√T) | Helped |
Closest papers by meaning
- Momentum-Based Variance Reduction in Non-Convex SGD (NeurIPS 2019 · similarity 0.94)
- Stability and Generalization for Stochastic Recursive Momentum-based Algorithms for (Strongly-)Convex One to $K$-Level Stochastic Optimizations (ICML 2024 · similarity 0.91)
- Adaptive Variance Reduction for Stochastic Optimization under Weaker Assumptions (NeurIPS 2024 · similarity 0.90)
- Random Scaling and Momentum for Non-smooth Non-convex Optimization (ICML 2024 · similarity 0.87)
- Better SGD using Second-order Momentum (NeurIPS 2022 · similarity 0.87)
- Proximal Stochastic Recursive Momentum Methods for Nonconvex Composite Decentralized Optimization (AAAI 2023 · similarity 0.87)
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.