Research map / Learning theory and algorithms / Matrix and tensor methods
Affine Rank Minimization via Asymptotic Log-Det Iteratively Reweighted Least Squares
Sebastian Krämer
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 |
|---|---|---|---|
| lower weight strength p from 1 to 0 (log-det limit) (ablation) | IRLS-p · Matrix recovery, Gaussian, l=63 | recovery rate (%): 0.0 -> 31.9 | Improved |
| lower weight strength p from 1 to 0 (ablation) | IRLS-p · Matrix recovery, Gaussian, l=72 | recovery rate (%): 0.0 -> 99.0 | Improved |
| lower weight strength p from 0.8 to 0 (ablation) | IRLS-p · Matrix recovery, Gaussian, l=72 | recovery rate (%): 49.8 -> 99.0 | Improved |
| lower weight strength p from 1.0 to 0 (ablation) | IRLS-p · Matrix recovery, Gaussian, l=108 | recovery rate (%): 98.8 -> 100.0 | No clear effect |
| use both-sided weights instead of one-sided (ablation) | IRLS-0 · Matrix recovery, Gaussian, l=63, kmax=14 | recovery rate (%): 34.0 -> 43.2 | Improved |
| use both-sided weights instead of one-sided (ablation) | IRLS-0 · Matrix recovery, sampling, l=64, kmax=14 | recovery rate (%): 22.9 -> 35.9 | Improved |
| use both-sided weights instead of one-sided (ablation) | IRLS-0 · Matrix recovery, sampling, l=63, kmax=12 | recovery rate (%): 5.6 -> 24.0 | Improved |
| increase measurements from 62 to 64 | IRLS-0 gaussian · Matrix recovery, kmax=14 | recovery rate (%): 0.0 -> 58.2 | Improved |
| increase measurements from 63 to 72 | IRLS-0 gaussian · Matrix recovery, kmax=12 | recovery rate (%): 31.9 -> 99.0 | Improved |
| use Gaussian measurements instead of sampling | IRLS-0 · Matrix recovery, l=72, kmax=12 | recovery rate (%): 77.1 -> 99.0 | Improved |
| increase measurements from 72 to 108 for sampling | IRLS-0 sampling · Matrix recovery, kmax=12 | recovery rate (%): 77.1 -> 100.0 | Improved |
Closest papers by meaning
- Harmonic Mean Iteratively Reweighted Least Squares for Low-Rank Matrix Recovery (JMLR 2018 · similarity 0.90)
- Recovery guarantee of weighted low-rank approximation via alternating minimization (ICML 2016 · similarity 0.90)
- Symmetric Matrix Completion with ReLU Sampling (ICML 2024 · similarity 0.88)
- Precise asymptotics of reweighted least-squares algorithms for linear diagonal networks (NeurIPS 2024 · similarity 0.87)
- Guarantees of a Preconditioned Subgradient Algorithm for Overparameterized Asymmetric Low-rank Matrix Recovery (ICML 2025 · similarity 0.87)
- Efficient Alternating Minimization with Applications to Weighted Low Rank Approximation (ICLR 2025 · similarity 0.87)
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.
