Research map / Learning theory and algorithms / Matrix and tensor methods
Optimal Shrinkage of Singular Values Under Random Data Contamination
Danny Barash, Matan Gavish
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 |
|---|---|---|---|
| Replace TSVD with the optimal singular value shrinker | SVD-based low-rank reconstruction · Asymptotic low-rank matrix model, β=1 | worst-case mean square error: (σ_B/μ_A)^2 5r -> (σ_B/μ_A)^2 2r | Improved |
| Replace TSVD with the optimal hard threshold | SVD-based low-rank reconstruction · Asymptotic low-rank matrix model, β=1 | worst-case mean square error: (σ_B/μ_A)^2 5r -> (σ_B/μ_A)^2 3r | Improved |
| Use the optimal singular value shrinker rather than the optimal hard threshold | SVD-based low-rank reconstruction · Asymptotic low-rank matrix model, β=1 | worst-case mean square error: (σ_B/μ_A)^2 3r -> (σ_B/μ_A)^2 2r | Improved |
| Use the optimal shrinker instead of the zero shrinker | Low-rank matrix reconstruction · Asymptotic low-rank matrix model | critical signal level: (σ_B / μ_A) · β^\frac{1}{4} | Improved |
Closest papers by meaning
- The Power of Iterative Filtering for Supervised Learning with (Heavy) Contamination (NeurIPS 2025 · similarity 0.87)
- Fast exact recovery of noisy matrix from few entries: the infinity norm approach (NeurIPS 2025 · similarity 0.86)
- Matrix Denoising with Doubly Heteroscedastic Noise: Fundamental Limits and Optimal Spectral Methods (NeurIPS 2024 · similarity 0.86)
- Robust Sparse Estimation for Gaussians with Optimal Error under Huber Contamination (ICML 2024 · similarity 0.85)
- Efficient Multivariate Robust Mean Estimation Under Mean-Shift Contamination (ICML 2025 · similarity 0.84)
- Guarantees of a Preconditioned Subgradient Algorithm for Overparameterized Asymmetric Low-rank Matrix Recovery (ICML 2025 · similarity 0.84)
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.
