Bayesian bandwidth test and selection for high-dimensional banded precision matrices

DOI10.1214/19-BA1167zbMATH Open1459.62086arXiv1804.08650OpenAlexW2950462788WikidataQ127641066 ScholiaQ127641066MaRDI QIDQ2226705FDOQ2226705

Authors: Kyoungjae Lee, Lizhen Lin 

Publication date: 9 February 2021

Published in: Bayesian Analysis (Search for Journal in Brave)

Abstract: Assuming a banded structure is one of the common practice in the estimation of high-dimensional precision matrix. In this case, estimating the bandwidth of the precision matrix is a crucial initial step for subsequent analysis. Although there exist some consistent frequentist tests for the bandwidth parameter, bandwidth selection consistency for precision matrices has not been established in a Bayesian framework. In this paper, we propose a prior distribution tailored to the bandwidth estimation of high-dimensional precision matrices. The banded structure is imposed via the Cholesky factor from the modified Cholesky decomposition. We establish the strong model selection consistency for the bandwidth as well as the consistency of the Bayes factor. The convergence rates for Bayes factors under both the null and alternative hypotheses are derived which yield similar order of rates. As a by-product, we also proposed an estimation procedure for the Cholesky factors yielding an almost optimal order of convergence rates. Two-sample bandwidth test is also considered, and it turns out that our method is able to consistently detect the equality of bandwidths between two precision matrices. The simulation study confirms that our method in general outperforms the existing frequentist and Bayesian methods.

Full work available at URL: https://arxiv.org/abs/1804.08650

Recommendations

• Test for bandedness of high-dimensional precision matrices
• Forward adaptive banding for estimating large covariance matrices
• Test for bandedness of high-dimensional covariance matrices and bandwidth estimation
• Hypothesis testing for band size detection of high-dimensional banded precision matrices
• Estimating large precision matrices via modified Cholesky decomposition

zbMATH Keywords

bandwidth selectionprecision matrixCholesky factorconvergence rates of Bayes factor

Mathematics Subject Classification ID

Asymptotic properties of parametric tests (62F05) Estimation in multivariate analysis (62H12) Hypothesis testing in multivariate analysis (62H15) Bayesian problems; characterization of Bayes procedures (62C10)

Cites Work

• Tractable Bayesian variable selection: beyond normality
• Empirical Bayes posterior concentration in sparse high-dimensional linear models
• Scalable Bayesian variable selection using nonlocal prior densities in ultrahigh-dimensional settings
• Asymptotic normality and optimalities in estimation of large Gaussian graphical models
• High dimensional covariance matrix estimation using a factor model
• Regularized estimation of large covariance matrices
• Mixtures of g Priors for Bayesian Variable Selection
• Two-Sample Covariance Matrix Testing and Support Recovery in High-Dimensional and Sparse Settings
• On consistency and sparsity for principal components analysis in high dimensions
• Posterior contraction in sparse Bayesian factor models for massive covariance matrices
• Bayesian model selection in high-dimensional settings
• Consistency of objective Bayes factors as the model dimension grows
• Cholesky decomposition of a hyper inverse Wishart matrix
• Minimax estimation of large covariance matrices under \(\ell_1\)-norm
• Covariance matrix selection and estimation via penalised normal likelihood
• Optimal rates of convergence for covariance matrix estimation
• Consistency of Bayesian linear model selection with a growing number of parameters
• Bayes factor consistency for nested linear models with a growing number of parameters
• A note on the consistency of Bayes factors for testing point null versus nonparametric alternatives.
• Bayesian linear regression with sparse priors
• On the use of Non-Local Prior Densities in Bayesian Hypothesis Tests
• Rate-optimal posterior contraction for sparse PCA
• Posterior convergence rates for estimating large precision matrices using graphical models
• On the computational complexity of high-dimensional Bayesian variable selection
• Optimal estimation and rank detection for sparse spiked covariance matrices
• Consistency of Bayes factor for nonnested model selection when the model dimension grows
• High dimensional posterior convergence rates for decomposable graphical models
• Posterior graph selection and estimation consistency for high-dimensional Bayesian DAG models
• Bayesian structure learning in graphical models
• Optimal Bayesian minimax rates for unconstrained large covariance matrices
• Minimax posterior convergence rates and model selection consistency in high-dimensional DAG models based on sparse Cholesky factors
• Hypothesis testing for band size detection of high-dimensional banded precision matrices
• Test for bandedness of high-dimensional precision matrices

Cited In (7)

• Test for bandedness of high-dimensional covariance matrices and bandwidth estimation
• Hypothesis testing for band size detection of high-dimensional banded precision matrices
• Estimating large precision matrices via modified Cholesky decomposition
• Test for bandedness of high-dimensional precision matrices
• Bandwidth selection for high-dimensional covariance matrix estimation
• Connection between the selection problem for a sparse submatrix of a large-size matrix and the Bayesian problem of hypotheses testing
• Scalable Bayesian high-dimensional local dependence learning

This page was built for publication: Bayesian bandwidth test and selection for high-dimensional banded precision matrices

Report a bug (only for logged in users!)Click here to report a bug for this page (MaRDI item Q2226705)