Research map / Optimization / Nonconvex and smooth optimization
A projected semismooth Newton method for a class of nonconvex composite programs with strong prox-regularity
Jiang Hu, Kangkang Deng, Jiayuan Wu, Quanzheng Li
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 ProxSSN instead of ProxGD | ProxSSN · Sparse PCA oblique (m/n/p=100/500/10) | wall-clock time (s): 13.59 -> 1.58 | Improved |
| use ProxSSN instead of AManPG | ProxSSN · Sparse PCA oblique (m/n/p=100/1000/30) | wall-clock time (s): 17.01 -> 5.37 | Improved |
| use ProxSSN instead of ARPG | ProxSSN · Sparse PCA oblique (m/n/p=100/1000/30) | wall-clock time (s): 15.79 -> 5.37 | Improved |
| use ProxSSN instead of ProxGD | ProxSSN · Sparse least squares (m/n=20/4600) | objective value: 3.71441e-02 -> 3.39844e-02 | Improved |
| use ProxSSN instead of ProxGD | ProxSSN · Sparse least squares (m/n=20/4600) | wall-clock time (s): 6.59 -> 0.13 | Improved |
| use ProxSSN instead of AManPG | ProxSSN · Sparse least squares (m/n=20/4600) | objective value: 3.35500e-02 -> 3.39844e-02 | Hurt |
| use ProxSSN instead of ProxGD | ProxSSN · Nonnegative PCA (n/p=600/25) | wall-clock time (s): 14.71 -> 0.90 | Improved |
| use ProxSSN instead of ProxGD | ProxSSN · Nonnegative PCA (n/p=800/15) | objective value: 1.839436 -> 1.837726 | Improved |
| use ProxSSN instead of ProxGD | ProxSSN · BEC (beta=500, Omega=0.00, initial a) | wall-clock time (s): 10.75 -> 0.17 | Improved |
| use ProxSSN instead of ARNT | ProxSSN · BEC (beta=500, Omega=0.25, initial a) | wall-clock time (s): 4.74 -> 1.14 | Improved |
Closest papers by meaning
- Skip the Hessian, Keep the Rates: Globalized Semismooth Newton with Lazy Hessian Updates (NeurIPS 2026 · similarity 0.87)
- A Regularized Newton Method for Nonconvex Optimization with Global and Local Complexity Guarantees (NeurIPS 2025 · similarity 0.86)
- Inexact Newton-type Methods for Optimisation with Nonnegativity Constraints (ICML 2024 · similarity 0.86)
- Zeroth-Order Methods for Constrained Nonconvex Nonsmooth Stochastic Optimization (ICML 2024 · similarity 0.85)
- Nonlinearly Preconditioned Gradient Methods under Generalized Smoothness (ICML 2025 · similarity 0.85)
- A Bregman Proximal Stochastic Gradient Method with Extrapolation for Nonconvex Nonsmooth Problems (AAAI 2024 · similarity 0.85)
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.
