Research map / Learning theory and algorithms / Matrix and tensor methods
Harmonic Mean Iteratively Reweighted Least Squares for Low-Rank Matrix Recovery
Christian Kümmerle, Juliane Sigl
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 harmonic mean weights (HM-IRLS) instead of arithmetic mean or column weights | IRLS for low-rank recovery · Matrix completion 100x100 rank-8 (150 instances) | empirical recovery probability: almost 100% at rho = 1.5 -> 100% from rho = 1.2 | Improved |
| use HM-IRLS instead of non-IRLS algorithms | low-rank recovery · Matrix completion 100x100 rank-8 (150 instances) | empirical recovery probability: 100% from rho = 1.2 vs never recover for rho < 1.2 | Improved |
| use small non-convex p (<= 0.25) in HM-IRLS (ablation) | HM-IRLS · Matrix completion rho=1.0 | relative Frobenius error: p <= 0.25 | Improved |
| use p=0.8 instead of small p (ablation) | HM-IRLS · Matrix completion rho=1.2 | relative Frobenius error: 0.27 after 100 iterations | Hurt |
| use IRLS-MF smoothing instead of IRLS-col smoothing | IRLS · Matrix completion 100x100 rank-8 (150 instances) | empirical recovery probability | Hurt |
Closest papers by meaning
- Recovering Simultaneously Structured Data via Non-Convex Iteratively Reweighted Least Squares (NeurIPS 2023 · similarity 0.93)
- Iteratively Reweighted Least Squares for Basis Pursuit with Global Linear Convergence Rate (NeurIPS 2021 · similarity 0.91)
- Precise asymptotics of reweighted least-squares algorithms for linear diagonal networks (NeurIPS 2024 · similarity 0.91)
- Guarantees of a Preconditioned Subgradient Algorithm for Overparameterized Asymmetric Low-rank Matrix Recovery (ICML 2025 · similarity 0.88)
- Fast exact recovery of noisy matrix from few entries: the infinity norm approach (NeurIPS 2025 · similarity 0.87)
- RGNMR: A Gauss-Newton method for robust matrix completion with theoretical guarantees (NeurIPS 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.
