An empirical Bayes procedure for the selection of Gaussian graphical models

DOI10.1007/S11222-011-9285-5zbMATH Open1272.62008arXiv1003.5851OpenAlexW2117101570WikidataQ60461473 ScholiaQ60461473MaRDI QIDQ693346FDOQ693346

Authors: Sophie Donnet, Jean-Michel Marin 

Publication date: 7 December 2012

Published in: Statistics and Computing (Search for Journal in Brave)

Abstract: A new methodology for model determination in decomposable graphical Gaussian models is developed. The Bayesian paradigm is used and, for each given graph, a hyper inverse Wishart prior distribution on the covariance matrix is considered. This prior distribution depends on hyper-parameters. It is well-known that the models's posterior distribution is sensitive to the specification of these hyper-parameters and no completely satisfactory method is registered. In order to avoid this problem, we suggest adopting an empirical Bayes strategy, that is a strategy for which the values of the hyper-parameters are determined using the data. Typically, the hyper-parameters are fixed to their maximum likelihood estimations. In order to calculate these maximum likelihood estimations, we suggest a Markov chain Monte Carlo version of the Stochastic Approximation EM algorithm. Moreover, we introduce a new sampling scheme in the space of graphs that improves the add and delete proposal of Armstrong et al. (2009). We illustrate the efficiency of this new scheme on simulated and real datasets.

Full work available at URL: https://arxiv.org/abs/1003.5851

Recommendations

• Decomposable graphical Gaussian model determination
• Bayesian graph selection consistency under model misspecification
• Objective Bayesian model selection in Gaussian graphical models
• Bayesian graphical Lasso models and efficient posterior computation
• scientific article; zbMATH DE number 5769726

zbMATH Keywords

Markov chain Monte CarloGaussian graphical modelsempirical Bayesdecomposable modelsstochastic approximation EM

Mathematics Subject Classification ID

Computational methods in Markov chains (60J22) Multivariate analysis (62H99) Empirical decision procedures; empirical Bayes procedures (62C12)

Cites Work

• Experiments in stochastic computation for high-dimensional graphical models
• Title not available (Why is that?)
• Hyper Markov laws in the statistical analysis of decomposable graphical models
• Title not available (Why is that?)
• The EM Algorithm and Extensions, 2E
• Model Selection and Accounting for Model Uncertainty in Graphical Models Using Occam's Window
• An application of MCMC methods for the multiple change-points problem.
• Bayes and empirical-Bayes multiplicity adjustment in the variable-selection problem
• Convergence of a stochastic approximation version of the EM algorithm
• Maximum likelihood estimation in nonlinear mixed effects models
• Wishart distributions for decomposable graphs
• Model selection for Gaussian concentration graphs
• Efficient Construction of Reversible Jump Markov Chain Monte Carlo Proposal Distributions
• Parametric Empirical Bayes Inference: Theory and Applications
• Simulation of hyper-inverse Wishart distributions in graphical models
• Limiting the Risk of Bayes and Empirical Bayes Estimators--Part II: The Empirical Bayes Case
• Multivariate empirical Bayes and estimation of covariance matrices
• Objective Bayesian model selection in Gaussian graphical models
• Hyper Inverse Wishart Distribution for Non-decomposable Graphs and its Application to Bayesian Inference for Gaussian Graphical Models
• Decomposable graphical Gaussian model determination
• Stein's Estimation Rule and Its Competitors--An Empirical Bayes Approach
• Empirical Bayes on vector observations: An extension of Stein's method
• Efficient estimation of covariance selection models
• Families of minimax estimators of the mean of a multivariate normal distribution
• Coupling a stochastic approximation version of EM with an MCMC procedure
• A stochastic approximation type EM algorithm for the mixture problem
• Asymptotic properties of a stochastic EM Algorithm for estimating mixing proportions
• Limiting the Risk of Bayes and Empirical Bayes Estimators--Part I: The Bayes Case
• Title not available (Why is that?)

Cited In (15)

• Bayesian graph selection consistency under model misspecification
• Decomposable graphical Gaussian model determination
• Efficient Gaussian graphical model determination under \(G\)-Wishart prior distributions
• Bayesian inference of clustering and multiple Gaussian graphical models selection
• A focused information criterion for graphical models
• Selecting the tuning parameter in penalized Gaussian graphical models
• Robust Bayesian graphical modeling using Dirichlet \(t\)-distributions
• Bayesian graphical model determination using decision theory
• Empirical Bayesian learning in AR graphical models
• Tuning parameter selection for penalized likelihood estimation of Gaussian graphical model
• An Expectation Conditional Maximization Approach for Gaussian Graphical Models
• Robust Bayesian model selection for variable clustering with the Gaussian graphical model
• Objective Bayesian model selection in Gaussian graphical models
• Title not available (Why is that?)
• Title not available (Why is that?)

Uses Software

• HdBCS

This page was built for publication: An empirical Bayes procedure for the selection of Gaussian graphical models

Report a bug (only for logged in users!)Click here to report a bug for this page (MaRDI item Q693346)