Research map / Learning theory and algorithms / Matrix and tensor methods
Nearly Optimal Robust Matrix Completion
Read the paperCheck 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 PG-RMC instead of St-NcRPCA | synthetic data at p=0.1 · Synthetic data | runtime to reach recovery error 0.01: ≈ 10s -> ≈ 2.5s | Helped |
| use PG-RMC instead of St-NcPCA | foreground-background separation · Shopping video | runtime: 204.4s -> 38.7s | Helped |
| use PG-RMC instead of St-NcPCA | foreground-background separation · Restaurant video | runtime: 27.8s -> 7.9s | Helped |
| use PG-RMC instead of Alternating Minimization | foreground-background separation · Escalator video | runtime: 150 seconds -> 8 seconds | Helped |
| reduce sampling probability to p=0.05 | PG-RMC foreground-background separation · Video datasets | background recovery quality | No clear effect |
| use rank-based stagewise updates, increasing rank by one per stage (ablation) | PG-RMC · RMC theoretical comparison | sample complexity: \order{\mu^{4}r^{2}n\log^2 \left(n\right) \log^2 \left(\frac{\sigma^{*}_{1}}{\epsilon}\right)} -> \order{\mu^{4}r^{3}n\l | Mixed |
| increase sampling probability p (ablation) | PG-RMC · Synthetic data | runtime | Mixed |
Closest papers by meaning
- RGNMR: A Gauss-Newton method for robust matrix completion with theoretical guarantees (NeurIPS 2025 · similarity 0.94)
- Fast Algorithms for Robust PCA via Gradient Descent (NeurIPS 2016 · similarity 0.93)
- Symmetric Matrix Completion with ReLU Sampling (ICML 2024 · similarity 0.91)
- Low Rank Matrix Completion via Robust Alternating Minimization in Nearly Linear Time (ICLR 2024 · similarity 0.90)
- Fast exact recovery of noisy matrix from few entries: the infinity norm approach (NeurIPS 2025 · similarity 0.89)
- Guarantees of a Preconditioned Subgradient Algorithm for Overparameterized Asymmetric Low-rank Matrix Recovery (ICML 2025 · 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.