Maximum Likelihood Estimation in Gaussian Chain Graph Models under the Alternative Markov Property

DOI10.1111/J.1467-9469.2006.00482.XzbMATH Open1125.62086arXivmath/0508266OpenAlexW2041282683WikidataQ57566459 ScholiaQ57566459MaRDI QIDQ3411073FDOQ3411073

Authors: Mathias Drton, Michael Eichler 

Publication date: 8 December 2006

Published in: Scandinavian Journal of Statistics (Search for Journal in Brave)

Abstract: The AMP Markov property is a recently proposed alternative Markov property for chain graphs. In the case of continuous variables with a joint multivariate Gaussian distribution, it is the AMP rather than the earlier introduced LWF Markov property that is coherent with data-generation by natural block-recursive regressions. In this paper, we show that maximum likelihood estimates in Gaussian AMP chain graph models can be obtained by combining generalized least squares and iterative proportional fitting to an iterative algorithm. In an appendix, we give useful convergence results for iterative partial maximization algorithms that apply in particular to the described algorithm.

Full work available at URL: https://arxiv.org/abs/math/0508266

Recommendations

• Iterative conditional fitting for discrete chain graph models.
• scientific article; zbMATH DE number 4211299
• Miscellanea. Maximum likelihood estimation in graphical models with missing values
• Gaussian Markov distributions over finite graphs
• Geometry of maximum likelihood estimation in Gaussian graphical models

zbMATH Keywords

efficiencyFisher informationiterative proportional fittingGaussian modelAnderson-Madigen-Perlman Markov chain graphical model

Mathematics Subject Classification ID

Markov processes: estimation; hidden Markov models (62M05) Markov processes (60J99)

Cites Work

• Introduction to Graphical Modelling
• Title not available (Why is that?)
• Title not available (Why is that?)
• Title not available (Why is that?)
• Ancestral graph Markov models.
• Model selection for Gaussian concentration graphs
• Title not available (Why is that?)
• A SINful approach to Gaussian graphical model selection
• Title not available (Why is that?)
• Alternative Markov properties for chain graphs
• Separation and completeness properties for AMP chain graph Markov models.
• Characterizing Markov equivalence classes for AMP chain graph models
• Multimodality of the likelihood in the bivariate seemingly unrelated regressions model
• Computing all roots of the likelihood equations of seemingly unrelated regressions

Cited In (14)

• Bayesian Structure Learning in Multilayered Genomic Networks
• Iterative conditional fitting for discrete chain graph models.
• The cost of using decomposable Gaussian graphical models for computational convenience
• Maximum likelihood for Gaussians on graphs
• Learning Markov equivalence classes of directed acyclic graphs: an objective Bayes approach
• Bayesian learning of graphical vector autoregressions with unequal lag-lengths
• Binary Models for Marginal Independence
• A SINful approach to Gaussian graphical model selection
• Multiple testing and error control in Gaussian graphical model selection
• Identifiability and Consistent Estimation for Gaussian Chain Graph Models
• Constrained likelihood for reconstructing a directed acyclic Gaussian graph
• Bayesian robust learning in chain graph models for integrative pharmacogenomics
• Characterizing Markov equivalence classes for AMP chain graph models
• Learning AMP chain graphs and some marginal models thereof under faithfulness

Uses Software

• MIM

This page was built for publication: Maximum Likelihood Estimation in Gaussian Chain Graph Models under the Alternative Markov Property

Report a bug (only for logged in users!)Click here to report a bug for this page (MaRDI item Q3411073)