Elliptical graphical modelling

DOI10.1093/BIOMET/ASR037MaRDI QIDQ3107982zbMATH OpenOpenAlexFDO

Authors Daniel Vogel, Roland Fried

Publication date 28 December 2011

Published in Biometrika (Search for Journal in Brave)

Full work available at URL https://arxiv.org/abs/1506.04321

zbMATH Keywords

concentration matrixdecomposable modeldeviance testpartial correlationTyler matrix

Mathematics Subject Classification ID

Applications of graph theory (05C90) Multivariate distribution of statistics (62H10) Estimation in multivariate analysis (62H12) Hypothesis testing in multivariate analysis (62H15)

Abstract: We propose elliptical graphical models based on conditional uncorrelatedness as a general- ization of Gaussian graphical models by letting the population distribution be elliptical instead of normal, allowing the fitting of data with arbitrarily heavy tails. We study the class of propor- tionally affine equivariant scatter estimators and show how they can be used to perform elliptical graphical modelling, leading to a new class of partial correlation estimators and analogues of the classical deviance test. General expressions for the asymptotic variance of partial correla- tion estimators, unconstrained and under decomposable models, are given, and the asymptotic chi square approximation of the pseudo-deviance test statistic is proved. The feasibility of our approach is demonstrated by a simulation study, using, among others, Tyler's scatter estimator, which is distribution-free within the elliptical model. Our approach provides a robustification of Gaussian graphical modelling. The latter is likelihood-based and known to be very sensitive to model misspecification and outlying observations.

Recommendations

• scientific article; zbMATH DE number 417563
• Elliptical multivariate analysis
• Multivariate elliptical models with general parameterization
• Predictive inference for the elliptical linear model
• scientific article; zbMATH DE number 1264885
• Graphical models
• Graphical models
• Gaussian graphical models

Cited in

(25)

• Estimation of high-dimensional graphical models using regularized score matching
• Robust methods for inferring sparse network structures
• Inference and mixture modeling with the elliptical Gamma distribution
• Optimal tests for elliptical symmetry: specified and unspecified location
• ROCKET: robust confidence intervals via Kendall's tau for transelliptical graphical models
• Dependence in elliptical partial correlation graphs
• A review of Gaussian Markov models for conditional independence
• Large covariance estimation through elliptical factor models
• Robust Gaussian graphical modeling
• Robust estimators for nondecomposable elliptical graphical models
• Robust and sparse estimation of the inverse covariance matrix using rank correlation measures
• Groups acting on Gaussian graphical models
• Predictive learning on hidden tree-structured Ising models
• Graphical lassos for meta‐elliptical distributions
• Nonparanormal graph quilting with applications to calcium imaging
• A nonparametric test for elliptical distribution based on kernel embedding of probabilities
• Transfer learning for high-dimensional transelliptical graphical models
• How robust are different versions of graphical model selection algorithms
• Asymptotic post-selection inference for regularized graphical models
• Non-independent component analysis
• Testing Elliptical Models in High Dimensions
• On some geometric identities involving the sample covariance matrix and its adjugate
• Pitman efficiency lower bounds for multivariate distribution-free tests based on optimal transport
• Elliptical regression operationalized
• Asymptotics of the two-stage spatial sign correlation

This page was built for publication: Elliptical graphical modelling

Report a bug (only for logged in users!)Click here to report a bug for this page (MaRDI item Q3107982)