Global identifiability of linear structural equation models
From MaRDI portal
Abstract: Structural equation models are multivariate statistical models that are defined by specifying noisy functional relationships among random variables. We consider the classical case of linear relationships and additive Gaussian noise terms. We give a necessary and sufficient condition for global identifiability of the model in terms of a mixed graph encoding the linear structural equations and the correlation structure of the error terms. Global identifiability is understood to mean injectivity of the parametrization of the model and is fundamental in particular for applicability of standard statistical methodology.
Recommendations
- Half-trek criterion for generic identifiability of linear structural equation models
- Identifiability of Gaussian structural equation models with equal error variances
- Generic identifiability of linear structural equation models by ancestor decomposition
- Identifiability of Gaussian linear structural equation models with homogeneous and heterogeneous error variances
- Algebraic problems in structural equation modeling
Cites work
- scientific article; zbMATH DE number 45532 (Why is no real title available?)
- scientific article; zbMATH DE number 964896 (Why is no real title available?)
- Ancestral graph Markov models.
- Asymptotic size and a problem with subsampling and with the \(m\) out of \(n\) bootstrap
- Causality. Models, reasoning, and inference
- Causation, prediction, and search
- Computing maximum likelihood estimates in recursive linear models with correlated errors
- Distinctness of the eigenvalues of a quadratic form in a multivariate sample
- Ideals, varieties, and algorithms. An introduction to computational algebraic geometry and commutative algebra
- Likelihood ratio tests and singularities
- On a parametrization of positive semidefinite matrices with zeros
- Probability distributions with summary graph structure
- What can we learn from the path equations?: identifiability, constraints, equivalence
Cited in
(30)- Algebraic problems in structural equation modeling
- Positivity for Gaussian graphical models
- A survey of discrete methods in (algebraic) statistics for networks
- Global identification of the semiparametric Box-Cox model
- Beyond the mean: a flexible framework for studying causal effects using linear models
- Identifying dynamical time series model parameters from equilibrium samples, with application to gene regulatory networks
- On the causal interpretation of acyclic mixed graphs under multivariate normality
- Nested covariance determinants and restricted trek separation in Gaussian graphical models
- Identification of discrete concentration graph models with one hidden binary variable
- Identifiability and equivalence of GLLIRM models
- Symmetries in directed Gaussian graphical models
- Algebraic properties of HTC-identifiable graphs
- Convergence rates of latent topic models under relaxed identifiability conditions
- What can we learn from the path equations?: identifiability, constraints, equivalence
- Complexity analysis of Bayesian learning of high-dimensional DAG models and their equivalence classes
- Counterfactual analyses with graphical models based on local independence
- Global identification of linearized DSGE models
- On the intrinsic relations of correlations in some systems of linear structural equations
- SEMID
- Causal inference in partially linear structural equation models
- Identifiability of nonrecursive structural equation models
- Comments on: Sequences of regressions and their independencies
- Identifiability of Gaussian structural equation models with equal error variances
- Generic global identification in factor analysis
- Identifiability in Continuous Lyapunov Models
- Dynamic structural equation models for directed cyclic graphs: the structural identifiability problem
- Identifiability of Gaussian linear structural equation models with homogeneous and heterogeneous error variances
- Generic identifiability of linear structural equation models by ancestor decomposition
- Half-trek criterion for generic identifiability of linear structural equation models
- Identifiability of directed Gaussian graphical models with one latent source
This page was built for publication: Global identifiability of linear structural equation models
Report a bug (only for logged in users!)Click here to report a bug for this page (MaRDI item Q116503)
