Research map / Learning theory and algorithms / Matrix and tensor methods
Recovery Guarantee of Non-negative Matrix Factorization via Alternating Updates
Yuanzhi Li, Yingyu Liang, Andrej Risteski
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 |
|---|---|---|---|
| Add ReLU thresholding (offset alpha) to the decoding step of the alternating update | alternating decode-update NMF algorithm (purification) · theoretical analysis, simplified sparse weights with adversarial noise | recovery error of feature matrix | Improved |
| Use unbiased zero-mean noise instead of adversarial noise (N2 vs N1) | alternating decode-update NMF algorithm · theoretical analysis, simplified case | maximum tolerable noise level relative to adversarial case: noise level sqrt(n) larger than adversarial case | Improved |
| Use minimum-infinity-norm pseudo-inverse for decoding rather than an arbitrary pseudo-inverse | decoding step · theoretical analysis | robustness to noise | Improved |
| Pre-processing step that rebalances feature scales (D scaling) to relax the xsu <= 2 xsl proportion restriction | general-proportion setting · theoretical analysis | applicability of guarantee | No clear effect |
| Use max_i row-wise potential instead of coupled potential a+beta*b for analysis | convergence analysis · theoretical analysis | potential function monotonicity | No clear effect |
Closest papers by meaning
- Provable Alternating Gradient Descent for Non-negative Matrix Factorization with Strong Correlations (ICML 2017 · similarity 0.91)
- Characterizing the Loss Landscape in Non-Negative Matrix Factorization (AAAI 2021 · similarity 0.88)
- Truncated Cauchy Non-Negative Matrix Factorization (TPAMI 2017 · similarity 0.87)
- Adversarial Nonnegative Matrix Factorization (ICML 2020 · similarity 0.87)
- An Exterior Method for Nonnegative Matrix Factorization (ICML 2026 · similarity 0.87)
- Non-negative Contrastive Learning (ICLR 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.
