Groups acting on Gaussian graphical models

DOI10.1214/13-AOS1130zbMATH Open1292.62098arXiv1207.5910OpenAlexW3103675751MaRDI QIDQ385776FDOQ385776

Authors: Piotr Zwiernik, Jan Draisma, S. Kuhnt 

Publication date: 11 December 2013

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

Abstract: Gaussian graphical models have become a well-recognized tool for the analysis of conditional independencies within a set of continuous random variables. From an inferential point of view, it is important to realize that they are composite exponential transformation families. We reveal this structure by explicitly describing, for any undirected graph, the (maximal) matrix group acting on the space of concentration matrices in the model. The continuous part of this group is captured by a poset naturally associated to the graph, while automorphisms of the graph account for the discrete part of the group. We compute the dimension of the space of orbits of this group on concentration matrices, in terms of the combinatorics of the graph; and for dimension zero we recover the characterization by Letac and Massam of models that are transformation families. Furthermore, we describe the maximal invariant of this group on the sample space, and we give a sharp lower bound on the sample size needed for the existence of equivariant estimators of the concentration matrix. Finally, we address the issue of robustness of these estimators by computing upper bounds on finite sample breakdown points.

Full work available at URL: https://arxiv.org/abs/1207.5910

Recommendations

• Automorphism groups of Gaussian Bayesian networks
• Graphical Gaussian Models With Edge and Vertex Symmetries
• Invariant normal models with recursive graphical Markov structure.
• On robust Gaussian graphical modeling
• Estimation of means in graphical Gaussian models with symmetries

zbMATH Keywords

Gaussian graphical modelscovariance matrixbreakdown pointequivariant estimatorrobust estimatorconcentration matrixtransformation families

Mathematics Subject Classification ID

Multivariate analysis (62H99) Robustness and adaptive procedures (parametric inference) (62F35) Transformation groups and semigroups (topological aspects) (54H15)

Cites Work

• Title not available (Why is that?)
• On a formula for the distribution of the maximum likelihood estimator
• Title not available (Why is that?)
• Testing Statistical Hypotheses
• Title not available (Why is that?)
• Title not available (Why is that?)
• Robust Statistics
• A General Qualitative Definition of Robustness
• Robust graphical modeling of gene networks using classical and alternative \(t\)-distributions
• Breakdown points of affine equivariant estimators of multivariate location and covariance matrices
• Two new properties of mathematical likelihood
• Robust Gaussian graphical modeling
• Breakdown and groups. (With discussions and rejoinder)
• Wishart distributions for decomposable graphs
• Elliptical graphical modelling
• Estimation with quadratic loss.
• The roles of conditioning in inference. With comments and rejoinder
• Title not available (Why is that?)
• Theory of statistics
• The real positive definite completion problem: cycle completability
• Lattice models for conditional independence in a multivariate normal distribution
• On the relation between conditional independence models determined by finite distributive lattices and by directed acyclic graphs
• On Riesz and Wishart distributions associated with decomposable undirected graphs
• Robust concentration graph model selection
• Estimation of the multivariate normal precision and covariance matrices in a star-shape model
• On the impact of contaminations in graphical Gaussian models
• Graphical methods for efficient likelihood inference in Gaussian covariance models
• Title not available (Why is that?)
• Groups acting on Gaussian graphical models
• Exponential transformation models
• Title not available (Why is that?)
• Geometry of maximum likelihood estimation in Gaussian graphical models
• The breakdown point -- examples and counterexamples
• Inadmissibility of the maximum likelihood estimator of normal covariance matrices with the lattice conditional independence

Cited In (9)

• Invariant normal models with recursive graphical Markov structure.
• Invariant Theory and Scaling Algorithms for Maximum Likelihood Estimation
• On the grouping effect of the \(l_{1-2}\) models
• Perturbation method for determining the group of invariance of hierarchical models
• Nonlinear algebra and applications
• A bridge between invariant theory and maximum likelihood estimation
• A Theory of Bayesian Groups
• Automorphism groups of Gaussian Bayesian networks
• Groups acting on Gaussian graphical models

Uses Software

• robustbase

This page was built for publication: Groups acting on Gaussian graphical models

Report a bug (only for logged in users!)Click here to report a bug for this page (MaRDI item Q385776)