High-dimensional covariance estimation based on Gaussian graphical models

Abstract

Undirected graphs are often used to describe high dimensional distributions. Under sparsity conditions, the graph can be estimated using ${\ell }_1$-penalization methods. We propose and study the following method. We combine a multiple regression approach with ideas of thresholding and refitting: first we infer a sparse undirected graphical model structure via thresholding of each among many ${\ell }_1$-norm penalized regression functions; we then estimate the covariance matrix and its inverse using the maximum likelihood estimator. We show that under suitable conditions, this approach yields consistent estimation in terms of graphical structure and fast convergence rates with respect to the operator and Frobenius norm for the covariance matrix and its inverse. We also derive an explicit bound for the Kullback Leibler divergence.

50 Pages, 6 figures. Major revision

References in corpus (7)

• Geometry of maximum likelihood estimation in Gaussian graphical models
• Adaptive Lasso for High Dimensional Regression and Gaussian Graphical Modeling
• Restricted Eigenvalue Conditions on Subgaussian Random Matrices
• Discussion: One-step sparse estimates in nonconcave penalized likelihood models
• High-dimensional covariance estimation by minimizing ${\ell }_1$-penalized log-determinant divergence
• Thresholded Lasso for high dimensional variable selection and statistical estimation
• The adaptive and the thresholded Lasso for potentially misspecified models

Cited by in corpus (22)

• Regularized rank-based estimation of high-dimensional nonparanormal graphical models
• Exact covariance thresholding into connected components for large-scale Graphical Lasso
• Gemini: Graph estimation with matrix variate normal instances
• Network Inference from Consensus Dynamics with Unknown Parameters
• Inferring gene-gene interactions and functional modules using sparse canonical correlation analysis
• Graph Learning from Data under Structural and Laplacian Constraints
• Reconstruction from anisotropic random measurements
• Inverse Covariance Estimation for High-Dimensional Data in Linear Time and Space: Spectral Methods for Riccati and Sparse Models
• The G-Wishart Weighted Proposal Algorithm: Efficient Posterior Computation for Gaussian Graphical Models
• Bayesian Learning of Graph Substructures
• Joint mean and covariance estimation with unreplicated matrix-variate data
• Graphical Elastic Net and Target Matrices: Fast Algorithms and Software for Sparse Precision Matrix Estimation
• High-dimensional Precision Matrix Estimation with a Known Graphical Structure
• Effective Learning of a GMRF Mixture Model
• Learning Gaussian Graphical Models via Multiplicative Weights
• A Stepwise Approach for High-Dimensional Gaussian Graphical Models
• The Sylvester Graphical Lasso (SyGlasso)
• Network-Based Pathway Enrichment Analysis with Incomplete Network Information
• Neighborhood selection with application to social networks
• Tensor models for linguistics pitch curve data of native speakers of Afrikaans
• Gaussian Graphical Model exploration and selection in high dimension low sample size setting
• Fine-Gray competing risks model with high-dimensional covariates: estimation and Inference