Wishart distributions for decomposable graphs

DOI10.1214/009053606000001235zbMATH Open1194.62078arXiv0708.2380OpenAlexW3106284700MaRDI QIDQ2642749FDOQ2642749

Authors: Gérard Letac, Hélène Massam 

Publication date: 4 September 2007

Published in: The Annals of Statistics (Search for Journal in Brave)

Abstract: When considering a graphical Gaussian model ${mathcalN}_G$ Markov with respect to a decomposable graph $G$, the parameter space of interest for the precision parameter is the cone $P_G$ of positive definite matrices with fixed zeros corresponding to the missing edges of $G$. The parameter space for the scale parameter of ${mathcalN}_G$ is the cone $Q_G$, dual to $P_G$, of incomplete matrices with submatrices corresponding to the cliques of $G$ being positive definite. In this paper we construct on the cones $Q_G$ and $P_G$ two families of Wishart distributions, namely the Type I and Type II Wisharts. They can be viewed as generalizations of the hyper Wishart and the inverse of the hyper inverse Wishart as defined by Dawid and Lauritzen [Ann. Statist. 21 (1993) 1272--1317]. We show that the Type I and II Wisharts have properties similar to those of the hyper and hyper inverse Wishart. Indeed, the inverse of the Type II Wishart forms a conjugate family of priors for the covariance parameter of the graphical Gaussian model and is strong directed hyper Markov for every direction given to the graph by a perfect order of its cliques, while the Type I Wishart is weak hyper Markov. Moreover, the inverse Type II Wishart as a conjugate family presents the advantage of having a multidimensional shape parameter, thus offering flexibility for the choice of a prior.

Full work available at URL: https://arxiv.org/abs/0708.2380

Recommendations

• Wishart distributions for decomposable covariance graph models
• On Riesz and Wishart distributions associated with decomposable undirected graphs
• Parameter priors for directed acyclic graphical models and the characterization of several probability distributions
• Exact formulas for the normalizing constants of Wishart distributions for graphical models
• Hyper Inverse Wishart Distribution for Non-decomposable Graphs and its Application to Bayesian Inference for Gaussian Graphical Models

zbMATH Keywords

graphical modelsnatural exponential familieschordal graphsconjugate priorshomogeneous conestriangulated graphsperfect ordercovariance selection modelsinverse Wisharthyper inverse Wisharthyper Markovhyper Wishart

Mathematics Subject Classification ID

Multivariate analysis (62H99) Applications of graph theory (05C90) Characterization and structure theory for multivariate probability distributions; copulas (62H05)

Cites Work

• Title not available (Why is that?)
• Title not available (Why is that?)
• Conjugate priors for exponential families
• Hyper Markov laws in the statistical analysis of decomposable graphical models
• Title not available (Why is that?)
• Title not available (Why is that?)
• Cholesky decomposition of a hyper inverse Wishart matrix
• Positive definite completions of partial Hermitian matrices
• On transformations and determinants of Wishart variables on symmetric cones
• Multivariate Beta Distributions and Independence Properties of the Wishart Distribution
• Wishart distributions on homogeneous cones
• Enriched conjugate and reference priors for the Wishart family on symmetric cones
• Title not available (Why is that?)
• Conditionally reducible natural exponential families and enriched conjugate priors
• Extremal rays and duals for cones of positive definite matrices with prescribed zeros

Cited In (63)

• Minimax estimation for mixtures of Wishart distributions
• Bayesian graph selection consistency under model misspecification
• Compatible priors for model selection of high-dimensional Gaussian DAGs
• Flexible covariance estimation in graphical Gaussian models
• Hierarchical Gaussian graphical models: beyond reversible jump
• Wishart laws and variance function on homogeneous cones
• An empirical Bayes procedure for the selection of Gaussian graphical models
• Graph-based multivariate conditional autoregressive models
• Linear optimization over homogeneous matrix cones
• Sparse matrix decompositions and graph characterizations
• Invariant Theory and Scaling Algorithms for Maximum Likelihood Estimation
• Posterior graph selection and estimation consistency for high-dimensional Bayesian DAG models
• Modeling correlated marker effects in genome-wide prediction via Gaussian concentration graph models
• Objective priors for generative star-shape models
• Reference priors for linear models with general covariance structures
• A characterization of the bivariate Wishart distribution
• Wigner and Wishart ensembles for sparse Vinberg models
• Characterization of the Riesz exponential family on homogeneous cones
• Wishart distributions for decomposable covariance graph models
• Functional equations of zeta functions associated with homogeneous cones
• The inverse Riesz probability distribution on symmetric matrices
• The G-Wishart Weighted Proposal Algorithm: Efficient Posterior Computation for Gaussian Graphical Models
• On Riesz and Wishart distributions associated with decomposable undirected graphs
• Post-processing posteriors over precision matrices to produce sparse graph estimates
• Accelerating Bayesian Structure Learning in Sparse Gaussian Graphical Models
• Riesz measures and Wishart laws associated to quadratic maps
• Posterior convergence rates for estimating large precision matrices using graphical models
• Determining full conditional independence by low-order conditioning
• Simultaneous predictive Gaussian classifiers
• A skew Gaussian decomposable graphical model
• Gaussian graphical modeling for spectrometric data analysis
• Parameter priors for directed acyclic graphical models and the characterization of several probability distributions
• Comment on: Sequences of regressions and their independences
• Marginal densities of the Wishart distribution.
• Extension of the Olkin and Rubin characterization to the Wishart distribution on homogeneous cones
• Improving on the maximum likelihood estimators of the means in Poisson decomposable graphical models
• The Inverse G‐Wishart distribution and variational message passing
• High dimensional posterior convergence rates for decomposable graphical models
• Bayesian inference in nonparanormal graphical models
• Statistical paleoclimate reconstructions via Markov random fields
• Bayesian structure learning in sparse Gaussian graphical models
• An accurate test for the equality of covariance matrices from decomposable graphical Gaussian models
• On the Letac-Massam conjecture and existence of high dimensional Bayes estimators for graphical models
• Exact formulas for the normalizing constants of Wishart distributions for graphical models
• Objective Bayes factors for Gaussian directed acyclic graphical models
• Building hyper Dirichlet processes for graphical models
• Conjugate and conditional conjugate Bayesian analysis of discrete graphical models of marginal independence
• A review of Gaussian Markov models for conditional independence
• Copula Gaussian graphical models and their application to modeling functional disability data
• On Gaussian group convex models
• On generating random Gaussian graphical models
• Wishart exponential families on cones related to tridiagonal matrices
• Bayesian structure learning in graphical models
• Bayesian hypothesis testing for Gaussian graphical models: conditional independence and order constraints
• Bayesian analysis of nonparanormal graphical models using rank-likelihood
• Groups acting on Gaussian graphical models
• Sparse covariance estimation in heterogeneous samples
• High-dimensional covariance estimation for Gaussian directed acyclic graph models with given order
• Gaussian graphical models with applications to omics analyses
• A Gibbs sampler for learning DAG: a unification for discrete and Gaussian domains
• Structural Markov graph laws for Bayesian model uncertainty
• A direct sampler for G-Wishart variates
• Posterior convergence rates for high-dimensional precision matrix estimation using \(G\)-Wishart priors

This page was built for publication: Wishart distributions for decomposable graphs

Report a bug (only for logged in users!)Click here to report a bug for this page (MaRDI item Q2642749)