Research map / Probabilistic and scientific modeling / Variational and generative sampling
Targeted Fused Ridge Estimation of Inverse Covariance Matrices from Multiple High-Dimensional Data Classes
Anders Ellern Bilgrau, Carel F.W. Peeters, Poul Svante Eriksen, Martin Boegsted, Wessel N. van Wieringen
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 unrestricted fused penalty matrix (LOOCV-selected) instead of the better of the non-fused ridge estimators | fused ridge estimator · Scenario 1 (banded, p=100, G=2) | Frobenius and quadratic loss | Improved |
| allow class-specific ridge penalties (unrestricted) instead of a single shared penalty lambda11=lambda22 (ablation) | fused ridge estimator · Scenario 1b (different class precision matrices) | Frobenius and quadratic loss | Improved |
| use pooled-covariance non-fused ridge when class precision matrices are equal | non-fused ridge · Scenario 1 (equal class matrices) | loss | Improved |
| add informative spot-on target T=Omega instead of null target (ablation) | fused ridge estimator · Scenario 2 (banded, star, random graph; p=50) | Frobenius and quadratic loss | Improved |
| add scalar diagonal target alpha*I instead of null target (ablation) | fused ridge estimator · Scenario 2 (all topologies, n_g=25,50,100) | loss | Improved |
| add scalar diagonal target instead of null target (ablation) | fused ridge estimator · Scenario 2i (banded, 25 bands, Frobenius loss) | Frobenius loss | Hurt |
| use rags2ridges instead of RidgeFusion package (null-target case) | fused ridge · p=50, G=2, n1=n2=10 banded | estimation time: 25.4 -> 9.3 milliseconds (2.74 speed-up) | Improved |
Closest papers by meaning
- A Well-Conditioned and Sparse Estimation of Covariance and Inverse Covariance Matrices Using a Joint Penalty (JMLR 2016 · similarity 0.91)
- Joint Structural Estimation of Multiple Graphical Models (JMLR 2016 · similarity 0.88)
- Shared Subspace Models for Multi-Group Covariance Estimation (JMLR 2019 · similarity 0.87)
- Estimation and Inference for High Dimensional Generalized Linear Models: A Splitting and Smoothing Approach (JMLR 2021 · similarity 0.87)
- Graphical Lasso and Thresholding: Equivalence and Closed-form Solutions (JMLR 2019 · similarity 0.87)
- Simultaneous Clustering and Estimation of Heterogeneous Graphical Models (JMLR 2018 · 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.
