Research map / Optimization / Nonconvex and smooth optimization
Stochastic subgradient for composite convex optimization with functional constraints
Ion Necoara, Nitesh Kumar Singh
Read the paperCheck your idea against 137,217 papers
AI-extracted. The measured effects below were read by AI from the paper's full text (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 robust (rho>0) instead of nominal (rho=0) formulation | SSP sparse SVM · COVID CT test, class-independent cov, lambda=0.3 | test accuracy: 159, 64.1% -> 198, 79.8% | Improved |
| use robust instead of nominal formulation | SSP sparse SVM · COVID CT test, class-dependent cov, lambda=0.2 | test accuracy: 198, 79.8% -> 200, 80.6% | No clear effect |
| use robust instead of nominal formulation | SSP sparse SVM · COVID CT test, class-dependent cov, lambda=0.1 | test accuracy: 166, 66.9% -> 180, 72.5% | Improved |
| use robust instead of nominal formulation | SSP sparse SVM · COVID CT test, class-independent cov, lambda=0.2 | test accuracy: 183, 73.8% -> 200, 80.6% | Improved |
| use SSP-LS instead of Leventhal-Lewis algorithm | least-squares solver · Random least-squares m=9000,p=10^5,n=10^4, delta=beta=1.96 | epochs: 213 -> 42 | Improved |
| use SSP-LS instead of Leventhal-Lewis algorithm | least-squares solver · Random least-squares m=9000,p=11000,n=10^5 | epochs: 23 -> 19 | No clear effect |
| increase delta=beta from 0.96 to 1.96 (ablation) | SSP-LS · Random least-squares m=9000,p=9000,n=10^4 | epochs: 1688 -> 1028 | Improved |
| increase delta=beta from 0.96 to 1.96 (ablation) | Leventhal-Lewis · Random least-squares m=9000,p=9000,n=10^4 | epochs: 1700 -> 1662 | No clear effect |
| use SSP-LS instead of Leventhal-Lewis | least-squares solver · Netlib LP fit2p (3000x13525) | cpu time (s): 2188.3 -> 65.7 | Improved |
| use SSP-LS instead of Matlab lsqlin | least-squares solver · Netlib LP qap12 (3193x8856) | cpu time (s): 3998.4 -> 374.1 | Improved |
| use SSP-LS instead of Matlab lsqlin | least-squares solver · Netlib LP afiro (21x51) | cpu time (s): 0.09 -> 1.9 | Hurt |
| use SSP-LS instead of Matlab lsqlin | least-squares solver · Netlib LP osa_07 (1118x25067) | cpu time (s): 3437.1 -> 3169.8 | No clear effect |
Closest papers by meaning
- Stochastic Conditional Gradient Methods: From Convex Minimization to Submodular Maximization (JMLR 2020 · similarity 0.91)
- A Generic Acceleration Framework for Stochastic Composite Optimization (NeurIPS 2019 · similarity 0.90)
- Almost surely constrained convex optimization (ICML 2019 · similarity 0.90)
- Universal Gradient Methods for Stochastic Convex Optimization (ICML 2024 · similarity 0.89)
- Universality of AdaGrad Stepsizes for Stochastic Optimization: Inexact Oracle, Acceleration and Variance Reduction (NeurIPS 2024 · similarity 0.88)
- Last-Iterate Convergence of Single-Loop Stochastic Methods for Constrained Convex-Concave Minimax Problems (NeurIPS 2026 · 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.
