Research map / Optimization / Nonconvex and smooth optimization
Conditional gradient methods for stochastically constrained convex minimization
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 increasing-size minibatches with SPIDER variance reduction instead of the single-sample variance-reduction scheme | H-1SFW, yielding H-SPIDER-FW · Stochastically constrained convex minimization | lmo calls: \mathcal{O}(\epsilon^{-6}) -> \mathcal{O}(\epsilon^{-2}) | Helped |
| Use increasing-size minibatches with SPIDER variance reduction instead of the single-sample variance-reduction scheme | H-1SFW, yielding H-SPIDER-FW · Stochastically constrained convex minimization | stochastic first-order oracle calls: \mathcal{O}(\epsilon^{-6}) -> \mathcal{O}(\epsilon^{-4}) | Helped |
| Process linear constraints stochastically using H-SPIDER-FW rather than SHCGM's deterministic constraint handling | Conditional-gradient methods for stochastic constraints · Stochastically constrained convex minimization | lmo calls: \mathcal{O}(\epsilon^{-3}) -> \mathcal{O}(\epsilon^{-2}) | Helped |
| Process constraints stochastically with H-1SFW rather than using SHCGM | Conditional-gradient methods for stochastic constraints · Stochastically constrained convex minimization | sfo calls: \mathcal{O}(\epsilon^{-3}) -> \mathcal{O}(\epsilon^{-6}) | Hurt |
| Use H-1SFW or H-SPIDER-FW instead of SHCGM | Conditional-gradient optimization on synthetic SDPs · Synthetic SDPs | objective suboptimality | Helped |
| Use H-1SFW or H-SPIDER-FW instead of SHCGM | Conditional-gradient optimization on synthetic SDPs · Synthetic SDPs | feasibility | Helped |
| Use H-1SFW or H-SPIDER-FW instead of SHCGM | K-means SDP relaxation · MNIST, k-means SDP | objective suboptimality | Helped |
| Use H-1SFW or H-SPIDER-FW instead of SHCGM | K-means SDP relaxation · MNIST, k-means SDP | feasibility | Helped |
| Use H-1SFW or H-SPIDER-FW instead of SHCGM | Uniform Sparsest Cut SDP · Network Repository graphs | objective suboptimality | Helped |
| Use H-1SFW or H-SPIDER-FW instead of SHCGM | Uniform Sparsest Cut SDP · Network Repository graphs | feasibility | Helped |
| Use the heavy-tailed Stable distribution rather than the benign synthetic-data distribution | H-1SFW and H-SPIDER-FW compared with SHCGM · Synthetic SDPs with heavy-tailed Stable entries | convergence speed | Mixed |
| Replace projection-based steps with conditional-gradient steps using an LMO | Stochastic constrained convex optimization · Stochastically constrained convex minimization | stochastic first-order oracle calls: \mathcal{O}(\epsilon^{-2}) -> \mathcal{O}(\epsilon^{-4}) | Hurt |
Closest papers by meaning
- Universal Gradient Methods for Stochastic Convex Optimization (ICML 2024 · similarity 0.89)
- Conditional Gradient Methods with Standard LMO for Stochastic Simple Bilevel Optimization (NeurIPS 2025 · similarity 0.89)
- Last-Iterate Convergence of Single-Loop Stochastic Methods for Constrained Convex-Concave Minimax Problems (NeurIPS 2026 · similarity 0.89)
- Stochastic Smoothed Primal-Dual Algorithms for Nonconvex Optimization with Linear Inequality Constraints (ICML 2025 · similarity 0.88)
- Single-Loop Stochastic Algorithms for Difference of Max-Structured Weakly Convex Functions (NeurIPS 2024 · similarity 0.88)
- Projection-Free Variance Reduction Methods for Stochastic Constrained Multi-Level Compositional Optimization (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.