Research map / Learning theory and algorithms / Matrix and tensor methods
A Well-Conditioned and Sparse Estimation of Covariance and Inverse Covariance Matrices Using a Joint Penalty
Ashwini Maurya
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 joint eigenvalue-variance plus l1 penalty instead of graphical lasso in LDA | LDA classifier · Colon tumor classification (p=200 genes) | classification error (%): 9.8 -> 8.2 | Improved |
| use joint penalty inverse covariance in LDA instead of graphical lasso | LDA classifier · Colon tumor classification (p=50 genes) | classification error (%): 10.9 -> 9.9 | Improved |
| use covariance-based LDA instead of naive Bayes | JPEN LDA · Colon tumor classification (p=100 genes) | classification error (%): 14.33 -> 8.9 | Improved |
| use JPEN covariance estimate instead of Ledoit-Wolf shrinkage | covariance estimator · Cov-I covariance estimation (n=50, p=500) | average relative error (%): 33.2 -> 7.1 | Improved |
| use JPEN covariance estimate instead of graphical lasso | covariance estimator · Cov-I covariance estimation (n=100, p=1000) | average relative error (%): 14.9 -> 7.8 | Improved |
| use JPEN covariance estimate instead of Bickel-Levina thresholding | covariance estimator · Toeplitz covariance estimation (n=50, p=500) | average relative error (%): 0.858 -> 2.517 | Hurt |
| use JPEN covariance estimate instead of PDSCE | covariance estimator · Hub covariance estimation (n=50, p=500) | average relative error (%): 0.735 -> 0.732 | No clear effect |
| use JPEN inverse covariance estimate instead of CLIME | inverse covariance estimator · Hub inverse covariance estimation (n=50, p=500) | average relative error (%): 10.5 -> 0.684 | Improved |
| use JPEN inverse covariance estimate instead of graphical lasso | inverse covariance estimator · Block inverse covariance estimation (n=100, p=500) | average relative error (%): 0.168 -> 0.562 | Hurt |
| increase number of genes p in JPEN-based LDA | JPEN LDA · Colon tumor classification | classification error (%): 9.9 -> 8.2 | Mixed |
Closest papers by meaning
- Targeted Fused Ridge Estimation of Inverse Covariance Matrices from Multiple High-Dimensional Data Classes (JMLR 2020 · similarity 0.91)
- Large-Scale Sparse Inverse Covariance Estimation via Thresholding and Max-Det Matrix Completion (ICML 2018 · similarity 0.91)
- Sparse PCA via Covariance Thresholding (JMLR 2016 · similarity 0.89)
- Positive Definite Sparse Covariance Estimation via Dual Space Optimization (AAAI 2026 · similarity 0.88)
- Non-Asymptotic Analysis Of Data Augmentation For Precision Matrix Estimation (NeurIPS 2025 · similarity 0.86)
- Fast and Provable Algorithms for Sparse PCA with Improved Sample Complexity (ICML 2025 · 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.
