Research map / Optimization / Nonconvex and smooth optimization
Revisiting Projection-Free Optimization for Strongly Convex Constraint Sets
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 Primal Averaging instead of standard FW with a predefined learning rate | Quadratic-loss regression with an l2 constraint · Regression | runtime: 3.7× | Helped |
| Use Primal Averaging instead of standard FW with line search | Quadratic-loss regression with an l2 constraint · Regression | runtime: 15.6× | Helped |
| Use Primal Averaging instead of projected gradient descent | Quadratic-loss regression · Regression | runtime: 7.7× | Helped |
| Use SGD instead of Stochastic Primal Averaging | Stochastic optimization with an l2 constraint · Regression | runtime: 4.6× | Helped |
| Use Stochastic Primal Averaging instead of SGD | Stochastic optimization with an l1.1 constraint · Regression | runtime: 25.1× | Helped |
| Use Primal Averaging | Convex loss over a strongly convex constraint · Convex-loss regression and matrix completion | runtime: 3.7--15.6× for a regression task and 2.8--11.7× for a matrix completion task | Helped |
| Use Primal Averaging on a perturbed objective | Convex loss over a strongly convex constraint · Convex loss | convergence slope: -2.41 | Helped |
| Use FW with line search | Squared-sigmoid loss with an l2 constraint · Strictly-locally-quasi-convex classification | convergence slope: -2.12 | Helped |
| Use FW with line search | Bi-weight loss with an l2 constraint · Non-convex robust regression | convergence slope: -1.46 instead of -1 | Helped |
Closest papers by meaning
- Lower Bounds for Frank-Wolfe on Strongly Convex Sets (ICML 2026 · similarity 0.92)
- A Richer Theory of Convex Constrained Optimization with Reduced Projections and Improved Rates (ICML 2017 · similarity 0.92)
- Frank-Wolfe Beyond 1/t Convergence (NeurIPS 2026 · similarity 0.91)
- Revisiting Frank-Wolfe for Structured Nonconvex Optimization (NeurIPS 2025 · similarity 0.90)
- Faster Accelerated First-order Methods for Convex Optimization with Strongly Convex Function Constraints (NeurIPS 2024 · similarity 0.89)
- Sarah Frank-Wolfe: Methods for Constrained Optimization with Best Rates and Practical Features (ICML 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.