Research map / Optimization / Nonconvex and smooth optimization
A Stochastic Composite Gradient Method with Incremental Variance Reduction
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 incremental variance-reduced estimates for both the inner mapping and its Jacobian in composite optimization | CIVR · Composite expectation and finite-sum nonconvex problems | sample complexity: \cO\bigl(\epsilon^{-3/2}\bigr); \cO\bigl(\min\{\epsilon^{-3/2},\,n^{1/2}\epsilon^{-1}\}\bigr) | Helped |
| Use adaptive epoch lengths and batch sizes instead of fixed settings | CIVR · Composite expectation, nonconvex | sample complexity: \tilde{\cO}\bigl(\epsilon^{-3/2}\bigr) | Helped |
| Use an adaptive schedule that transitions to full-batch epoch initialization | CIVR · Composite finite-sum, nonconvex | sample complexity: \tilde{\cO}\bigl(\min\bigl\{\sqrt{n}\epsilon^{-1},\epsilon^{-3/2}\bigr\}) | Helped |
| Periodically restart CIVR under gradient dominance | CIVR · Composite expectation, gradient dominant | sample complexity: \cO\bigl(\bigl(\nu\epsilon^{-1}\bigr)\log\epsilon^{-1}\bigr) | Helped |
| Periodically restart CIVR under gradient dominance | CIVR · Composite finite-sum, gradient dominant | sample complexity: \cO\bigl(\bigl(n+\frac{\nu\sqrt{n}}{\eta}\bigr)\ln\frac{1}{\epsilon}\bigr) | Helped |
Closest papers by meaning
- A Composite Randomized Incremental Gradient Method (ICML 2019 · similarity 0.92)
- Stochastic Optimization with Variance Reduction for Infinite Datasets with Finite Sum Structure (NeurIPS 2017 · similarity 0.90)
- Stochastic Nested Variance Reduction for Nonconvex Optimization (NeurIPS 2018 · similarity 0.89)
- Stochastic Variance Reduction for Nonconvex Optimization (ICML 2016 · similarity 0.89)
- Double Variance Reduction: A Smoothing Trick for Composite Optimization Problems without First-Order Gradient (ICML 2024 · similarity 0.88)
- Efficient Continual Finite-Sum Minimization (ICLR 2024 · 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.