Identifiability of Gaussian linear structural equation models with homogeneous and heterogeneous error variances
From MaRDI portal
Publication:2131903
DOI10.1007/S42952-019-00019-7zbMATH Open1485.62073arXiv1901.10134OpenAlexW2997194834MaRDI QIDQ2131903FDOQ2131903
Authors: Gunwoong Park, Youngwhan Kim 
Publication date: 27 April 2022
Published in: Journal of the Korean Statistical Society (Search for Journal in Brave)
Abstract: In this work, we consider the identifiability assumption of Gaussian linear structural equation models (SEMs) in which each variable is determined by a linear function of its parents plus normally distributed error. It has been shown that linear Gaussian structural equation models are fully identifiable if all error variances are the same or known. Hence, this work proves the identifiability of Gaussian SEMs with both homogeneous and heterogeneous unknown error variances. Our new identifiability assumption exploits not only error variances, but edge weights; hence, it is strictly milder than prior work on the identifiability result. We further provide a structure learning algorithm that is statistically consistent and computationally feasible, based on our new assumption. The proposed algorithm assumes that all relevant variables are observed, while it does not assume causal minimality and faithfulness. We verify our theoretical findings through simulations and real multivariate data, and compare our algorithm to state-of-the-art PC, GES and GDS algorithms.
Full work available at URL: https://arxiv.org/abs/1901.10134
Recommendations
- Identifiability of Gaussian structural equation models with equal error variances
- Global identifiability of linear structural equation models
- Robust estimation of Gaussian linear structural equation models with equal error variances
- Identifiability in Multivariate Dynamic Linear Errors-in-Variables Models
- Learning high-dimensional Gaussian linear structural equation models with heterogeneous error variances
- Identifiability of covariance parameters in linear mixed effects models
- Identifiability in generalized linear models with random effects
- Identifiability of linear mixed effects models
- Identifiability of nonrecursive structural equation models
causal inferencestructural equation modelBayesian networkdirected acyclic graphical modelidentifiability
Computational methods for problems pertaining to statistics (62-08) Learning and adaptive systems in artificial intelligence (68T05) Probabilistic graphical models (62H22)
Cites Work
- Causation, prediction, and search
- Introduction to Graphical Modelling
- Learning Bayesian networks: The combination of knowledge and statistical data
- Estimating high-dimensional directed acyclic graphs with the PC-algorithm
- Title not available (Why is that?)
- Title not available (Why is that?)
- The max-min hill-climbing Bayesian network structure learning algorithm
- Title not available (Why is that?)
- Geometry of the faithfulness assumption in causal inference
- Identifiability of Gaussian structural equation models with equal error variances
- A linear non-Gaussian acyclic model for causal discovery
- 10.1162/153244303321897717
- Title not available (Why is that?)
- DirectLiNGAM: a direct method for learning a linear non-Gaussian structural equation model
- High-dimensional learning of linear causal networks via inverse covariance estimation
- CAM: causal additive models, high-dimensional order search and penalized regression
- The three faces of faithfulness
- Learning quadratic variance function (QVF) DAG models via overdispersion scoring (ODS)
- Title not available (Why is that?)
Cited In (14)
- Identifiability of Gaussian structural equation models with equal error variances
- Learning high-dimensional Gaussian linear structural equation models with heterogeneous error variances
- Generic identifiability of linear structural equation models by ancestor decomposition
- Identifiability of additive noise models using conditional variances
- Identifiability of covariance parameters in linear mixed effects models
- Computationally Efficient Learning of Gaussian Linear Structural Equation Models with Equal Error Variances
- Title not available (Why is that?)
- Exact variance formula for the estimated mean outcome with external intervention based on the front-door criterion in Gaussian linear structural equation models
- Robust estimation of Gaussian linear structural equation models with equal error variances
- Global identifiability of linear structural equation models
- Half-trek criterion for generic identifiability of linear structural equation models
- Densely connected sub-Gaussian linear structural equation model learning via \(\ell_1\)- and \(\ell_2\)-regularized regressions
- Scalable Bayesian high-dimensional local dependence learning
- Causal inference in partially linear structural equation models
Uses Software
This page was built for publication: Identifiability of Gaussian linear structural equation models with homogeneous and heterogeneous error variances
Report a bug (only for logged in users!)Click here to report a bug for this page (MaRDI item Q2131903)
