Research map / Optimization / Nonconvex and smooth optimization
Projection-Free Methods for Solving Nonconvex-Concave Saddle Point Problems
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 projection-free R-PDCG/CG-RPGA instead of projection-based AGP | R-PDCG, CG-RPGA · Robust multiclass classification (rcv1, news20) | convergence over time | Helped |
| projection onto nuclear norm ball in AGP | AGP · Robust multiclass classification | time per iteration: more than three iterations taking over 300 seconds | Hurt |
| use CG-RPGA (one-sided LMO) instead of R-PDCG | CG-RPGA · Dictionary learning | convergence rate | Helped |
| use LMO instead of projection for nuclear norm ball | R-PDCG, CG-RPGA · Dictionary learning | computation time | Helped |
| add strong concavity and set regularization mu = 0 (ablation) | Algorithm 2 (CG-RPGA) · Nonconvex-strongly concave | iteration complexity: O(ϵ−4) -> O(ϵ−2) | Helped |
| compare against SPFW | R-PDCG, CG-RPGA · Dictionary learning | convergence | Helped |
Closest papers by meaning
- Efficient Projection-free Algorithms for Saddle Point Problems (NeurIPS 2020 · similarity 0.93)
- Projection-Free Algorithms for Minimax Problems (ICML 2026 · similarity 0.93)
- Inexact Newton-type Methods for Optimisation with Nonnegativity Constraints (ICML 2024 · similarity 0.88)
- On Linear Convergence in Smooth Convex-Concave Bilinearly-Coupled Saddle-Point Optimization: Lower Bounds and Optimal Algorithms (ICML 2025 · similarity 0.86)
- Second-Order Min-Max Optimization with Lazy Hessians (ICLR 2025 · similarity 0.86)
- Private Algorithms for Stochastic Saddle Points and Variational Inequalities: Beyond Euclidean Geometry (NeurIPS 2024 · similarity 0.86)
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.