Smooth, identifiable supermodels of discrete DAG models with latent variables

DOI10.3150/17-BEJ1005MaRDI QIDQ1740511zbMATH OpenOpenAlexWikidataFDO

Authors Robin J. Evans, Thomas S. Richardson

Publication date 30 April 2019

Published in Bernoulli (Search for Journal in Brave)

Full work available at URL https://arxiv.org/abs/1511.06813, https://projecteuclid.org/euclid.bj/1551862837

zbMATH Keywords

Bayesian networkDAGnested Markov modelparameterization

Mathematics Subject Classification ID

Discrete-time Markov processes on general state spaces (60J05) Estimation in multivariate analysis (62H12) Neural nets and related approaches to inference from stochastic processes (62M45)

Abstract: We provide a parameterization of the discrete nested Markov model, which is a supermodel that approximates DAG models (Bayesian network models) with latent variables. Such models are widely used in causal inference and machine learning. We explicitly evaluate their dimension, show that they are curved exponential families of distributions, and fit them to data. The parameterization avoids the irregularities and unidentifiability of latent variable models. The parameters used are all fully identifiable and causally-interpretable quantities.

Recommendations

• Margins of discrete Bayesian networks
• Identification of discrete concentration graph models with one hidden binary variable
• Identifiability of parameters in latent structure models with many observed variables
• Causal Graphical Models with Latent Variables: Learning and Inference
• Identifiability of directed Gaussian graphical models with one latent source

Cites work

• A new approach to causal inference in mortality studies with a sustained exposure period—application to control of the healthy worker survivor effect
• Ancestral graph Markov models.
• Binary Models for Marginal Independence
• Causality. Models, reasoning, and inference
• Discrete chain graph models
• Influence Diagrams for Causal Modelling and Inference
• Likelihood ratio tests and singularities
• Marginal log-linear parameters for graphical Markov models
• Margins of discrete Bayesian networks
• Markov Properties for Acyclic Directed Mixed Graphs
• Markovian acyclic directed mixed graphs for discrete data
• Modeling and Reasoning with Bayesian Networks
• Nested Markov properties for acyclic directed mixed graphs
• Probability distributions with summary graph structure
• Stochastic factorizations, sandwiched simplices and the topology of the space of explanations
• The hidden life of latent variables: Bayesian learning with mixed graph models

Cited in

(3)

• Margins of discrete Bayesian networks
• Nested Markov properties for acyclic directed mixed graphs
• Conditional independence collapsibility for acyclic directed mixed graph models

Describes a project that uses

Uses Software

• PRMLT

This page was built for publication: Smooth, identifiable supermodels of discrete DAG models with latent variables

Report a bug (only for logged in users!)Click here to report a bug for this page (MaRDI item Q1740511)