Research map / Learning theory and algorithms / Matrix and tensor methods
Robust Formulation for PCA: Avoiding Mean Calculation With L2,p-norm Maximization
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 |
|---|---|---|---|
| smaller p (0.5) vs larger p (1.5) in l2,p-norm (ablation) | l2,p-PCA · COIL20 dim 10 | reconstruction error: 0.8457 -> 0.8169 | Helped |
| smaller p (0.5) vs larger p (1.5) in l2,p-norm (ablation) | l2,p-PCA · UMIST dim 10 | reconstruction error: 0.9087 -> 0.8873 | Helped |
| smaller p vs larger p, averaged over dimensions (ablation) | l2,p-PCA · ORL (avg) | reconstruction error: 0.8141 (p=0.5) vs 0.8069 (p=1.5) | No clear effect |
| smaller p vs larger p, averaged over dimensions (ablation) | l2,p-PCA · LFWCrop (avg) | reconstruction error: 0.5443 vs 0.5288 | No clear effect |
| avoid mean estimation via pairwise l2,p-norm instead of fixed-mean l1 RPCA | RPCA · ORL (avg) | reconstruction error: 0.8882 -> 0.8125 | Helped |
| pairwise l2,p-norm instead of pairwise l1-norm (RPCA-AOM) | RPCA-AOM · LFWCrop (avg) | reconstruction error: 0.6827 -> 0.5290 | Helped |
| l2,p-PCA vs RPCA-AOM runtime | RPCA-AOM · Coil20 | running time (minutes): 312.12 -> 17.81 | Helped |
Closest papers by meaning
- Nearly-Linear Time and Streaming Algorithms for Outlier-Robust PCA (ICML 2023 · similarity 0.91)
- A Pseudo-Bayesian Algorithm for Robust PCA (NeurIPS 2016 · similarity 0.89)
- Robust Principal Component Analysis with Side Information (ICML 2016 · similarity 0.89)
- Learned Robust PCA: A Scalable Deep Unfolding Approach for High-Dimensional Outlier Detection (NeurIPS 2021 · similarity 0.88)
- Residual-Based Sampling for Online Outlier-Robust PCA (ICML 2022 · similarity 0.88)
- Capturing the denoising effect of PCA via compression ratio (NeurIPS 2024 · similarity 0.86)
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.