Research map / Learning theory and algorithms / Matrix and tensor methods
Low-Rank Matrix Estimation in the Presence of Change-Points
Lei Shi, Guanghui Wang, Changliang Zou
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 |
|---|---|---|---|
| detect the change-point with matrix-structured nuclear-norm scheme instead of vectorized LASSO (Vec) | Ours · Compressed sensing (m=40, N=1500) | squared Frobenius error of Theta1: 0.937 -> 0.246 | Improved |
| use matrix-structured scheme instead of Vec for change-point localization | Ours · Multivariate regression (m=50, n=500) | change-point error: 0.040 -> 0.031 | Improved |
| account for the change-point instead of ignoring it (NC) | Ours · Compressed sensing (m=40, N=2500) | squared Frobenius error of Theta1: 0.700 -> 0.120 | Improved |
| account for the change-point instead of ignoring it (NC) | Ours · Multivariate regression (m=50, n=500) | squared Frobenius error of Theta1: 0.225 -> 0.352 | Hurt |
| estimate the change-point instead of knowing it (Oracle) | Ours · Multivariate regression (m=50, n=2000) | squared Frobenius error of Theta1: 0.107 -> 0.107 | No clear effect |
| increase sample size from 500 to 2000 | Ours · Multivariate regression (m=50) | change-point error: 0.031 -> 0.006 | Improved |
| increase sample size from 1500 to 2500 | Ours · Compressed sensing (m=40) | squared Frobenius error of Theta1: 0.246 -> 0.120 | Improved |
| use large instead of small break size | Ours · Multiple change-points | matrix overall error MOE: 0.291 -> 0.115 | Improved |
| add one change-point (0 to 1) | Ours · Beijing air pollution test set | test error: 0.1925 -> 0.1746 | Improved |
| add a second change-point (1 to 2) | Ours · Beijing air pollution test set | test error: 0.1746 -> 0.1728 | Improved |
| add a third change-point (2 to 3) | Ours · Beijing air pollution test set | test error: 0.1728 -> 0.1761 | Hurt |
Closest papers by meaning
- Guarantees of a Preconditioned Subgradient Algorithm for Overparameterized Asymmetric Low-rank Matrix Recovery (ICML 2025 · similarity 0.85)
- Accelerating Ill-Conditioned Low-Rank Matrix Estimation via Scaled Gradient Descent (JMLR 2021 · similarity 0.85)
- Robust Matrix Sensing in the Semi-Random Model (NeurIPS 2023 · similarity 0.84)
- Inferring Change Points in High-Dimensional Linear Regression via Approximate Message Passing (ICML 2024 · similarity 0.84)
- Efficient Low-Rank Matrix Estimation, Experimental Design, and Arm-Set-Dependent Low-Rank Bandits (ICML 2024 · similarity 0.83)
- A Subspace-Constrained Tyler's Estimator and its Applications to Structure from Motion (CVPR 2024 · similarity 0.83)
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.
