Research map / Learning theory and algorithms / Matrix and tensor methods
A Matrix Splitting Method for Composite Function Minimization
Check 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 matrix splitting instead of the compared NMF solvers | Nonnegative matrix factorization · 20news, COIL, and TDT2 | objective value: At 50 seconds and n=300: 20news MSM 2.331e+06 vs APG 3.621e+06; COIL MSM 4.127e+08 vs CGD 4.226e+08; TDT2 MSM 9.606e+05 | Helped |
| Use MSM rather than PGM to update sparse codes | Cardinality-regularized sparse coding · Image denoising | SNR: The text reports lower objectives in all cases; Table 2 shows higher SNR in most listed comparisons. | Helped |
| Use MSM rather than KSVD | Cardinality-regularized sparse coding · Image denoising | SNR: The paper reports comparable SNR and says MSM is generally 4-8 times faster than KSVD. | Mixed |
| Initialize sparse codes with OMP rather than random initialization | Cardinality-regularized sparse coding · Image denoising | SNR | No clear effect |
| Use MSM for the fixed-dictionary sparse-coding subproblem | Cardinality-regularized sparse coding · Sparse-coding objective with fixed DCT dictionary | objective value | Helped |
Closest papers by meaning
- An Alternating Proximal Splitting Method with Global Convergence for Nonconvex Structured Sparsity Optimization (AAAI 2016 · similarity 0.88)
- Polynomial Optimization Methods for Matrix Factorization (AAAI 2017 · similarity 0.86)
- Approximate Decomposable Submodular Function Minimization for Cardinality-Based Components (NeurIPS 2021 · similarity 0.86)
- GSOS: Gauss-Seidel Operator Splitting Algorithm for Multi-Term Nonsmooth Convex Composite Optimization (ICML 2017 · similarity 0.86)
- Convergence of Alternating Gradient Descent for Matrix Factorization (NeurIPS 2023 · similarity 0.86)
- Globally Q-linear Gauss-Newton Method for Overparameterized Non-convex Matrix Sensing (NeurIPS 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.