Estimating Hierarchically Rank Structured Covariance Matrices

DOI10.48550/arXiv.2609.09944MaRDI QIDQ7345532arXivFDO

Authors Robin Armstrong, Anil Damle, Samuel E. Otto

Publication date 9 September 2026

Published in arXiv

Tldr We introduce a hierarchical rank structure for estimating high-dimensional covariance matrices from small sample sizes. Our approach accommodates smooth long-range correlations while achieving efficient matrix-vector products through data-sparse representation.

Full work available at URL https://doi.org/10.48550/arXiv.2609.09944

arXiv classification

math.NAstat.CO

Cites work

• A fast direct solver for structured linear systems by recursive skeletonization
• A fast randomized algorithm for computing a hierarchically semiseparable representation of a matrix
• A recursive skeletonization factorization based on strong admissibility
• A well-conditioned estimator for large-dimensional covariance matrices
• Analytical nonlinear shrinkage of large-dimensional covariance matrices
• Covariance estimation: the GLM and regularization perspectives
• Covariance regularization by thresholding
• Data assimilation: methods, algorithms, and applications
• Fast algorithms for hierarchically semiseparable matrices
• Fast Direct Solvers for Elliptic PDEs
• Fast spatial Gaussian process maximum likelihood estimation via skeletonization factorizations
• Four lectures on probabilistic methods for data science
• Generalized thresholding of large covariance matrices
• Hierarchical interpolative factorization for elliptic operators: integral equations
• Hierarchical matrices: algorithms and analysis
• High-Dimensional Probability: An Introduction with Applications in Data Science
• Localized covariance estimation: A Bayesian perspective
• Matrix Analysis
• Nonparametric estimation of large covariance matrices of longitudinal data
• Orthogonal polynomials
• Randomized algorithms for low-rank matrix approximation: Design, analysis, and applications
• Regularized estimation of large covariance matrices
• Sparse inverse covariance estimation with the graphical lasso
• Surface quasi-geostrophic dynamics
• The Proper Orthogonal Decomposition in the Analysis of Turbulent Flows
• The Rotation of Eigenvectors by a Perturbation. III

This page was built for publication: Estimating Hierarchically Rank Structured Covariance Matrices

Report a bug (only for logged in users!)Click here to report a bug for this page (MaRDI item Q7345532)