Research map / Learning theory and algorithms / Matrix and tensor methods
Matrix Completion With Cross-Concentrated Sampling: Bridging Uniform Sampling and CUR Sampling
HanQin Cai, Longxiu Huang, Pengyu Li, Deanna M. Needell
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 CCS sampling with ICURC-8 instead of uniform sampling with ScaledPGD | matrix completion · Building image inpainting, alpha=10% | SNR (dB): 20.593 -> 23.762 | Improved |
| use CCS sampling with ICURC-8 instead of uniform sampling with SVP | matrix completion · Building image inpainting, alpha=10% | SNR (dB): 18.065 -> 23.762 | Improved |
| use ICURC-8 on CCS instead of ScaledPGD on uniform sampling | matrix completion · Building image inpainting, alpha=10% | runtime (sec): 3.518 -> 0.400 | Improved |
| use ICURC-8 on CCS instead of SVP on uniform sampling | matrix completion · Window image inpainting, alpha=10% | runtime (sec): 10.626 -> 0.339 | Improved |
| use ICURC-9 on CCS instead of ScaledPGD on uniform sampling | matrix completion · Window image inpainting, alpha=14% | SNR (dB): 32.693 -> 32.984 | Improved |
| change concentrated row/column submatrix size at fixed overall sampling rate (ICURC-8 vs ICURC-10) (ablation) | ICURC · Building image inpainting, alpha=14% | SNR (dB): 25.195 vs 24.823 | No clear effect |
| vary size of concentrated submatrices c in {0.25, 0.5, 1} (ablation) | ICURC on CCS · synthetic phase transition | required samples | No clear effect |
| increase matrix rank r from 5 to 15 (ablation) | ICURC on CCS · synthetic phase transition | required sampling rate | Hurt |
| use CCS with ICURC instead of uniform sampling with ScaledPGD and SVP | matrix completion · MovieLens-100K, MovieLens-10M, FilmTrust | hit rate and NMAE | Improved |
| use CCS with ICURC instead of uniform sampling with ScaledPGD and SVP | matrix completion · Blogs, Opsahl, Figeys link prediction | Precision and AUC | Improved |
Closest papers by meaning
- A New Theory for Matrix Completion (NeurIPS 2017 · similarity 0.89)
- Symmetric Matrix Completion with ReLU Sampling (ICML 2024 · similarity 0.89)
- Fast and Sample Efficient Inductive Matrix Completion via Multi-Phase Procrustes Flow (ICML 2018 · similarity 0.88)
- Matrix Completion with Deterministic Sampling: Theories and Methods (TPAMI 2019 · similarity 0.88)
- A Pairwise Pseudo-likelihood Approach for Matrix Completion with Informative Missingness (NeurIPS 2024 · similarity 0.87)
- Generalization Analysis of Deep Non-linear Matrix Completion (ICML 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.
