Consistent DAG selection for Bayesian causal discovery under general error distributions

Fuente: arXiv
Salvato in:
Dettagli Bibliografici
Autori principali: Chaudhuri, Anamitra, Bhattacharya, Anirban, Ni, Yang
Natura: Preprint
Pubblicazione: 2025
Soggetti:
Accesso online:
Tags: Aggiungi Tag
Nessun Tag, puoi essere il primo ad aggiungerne!!
_version_ 1866913971464830976
author Chaudhuri, Anamitra
Bhattacharya, Anirban
Ni, Yang
author_facet Chaudhuri, Anamitra
Bhattacharya, Anirban
Ni, Yang
contents We consider the problem of learning the underlying causal structure among a set of variables, which are assumed to follow a Bayesian network or, more specifically, a linear recursive structural equation model (SEM) with the associated errors being independent and allowed to be non-Gaussian. A Bayesian hierarchical model is proposed to identify the true data-generating directed acyclic graph (DAG) structure where the nodes and edges represent the variables and the direct causal effects, respectively. Moreover, incorporating the information of non-Gaussian errors, we characterize the distribution equivalence class of the true DAG, which specifies the best possible extent to which the DAG can be identified based on purely observational data. Furthermore, under the consideration that the errors are distributed as some scale mixture of Gaussian, where the mixing distribution is unspecified, and mild distributional assumptions, we establish that by employing a non-standard DAG prior, the posterior probability of the distribution equivalence class of the true DAG converges to unity as the sample size grows. This shows that the proposed method achieves the posterior DAG selection consistency, which is further illustrated with examples and simulation studies.
format Preprint
id arxiv_https___arxiv_org_abs_2508_00993
institution arXiv
publishDate 2025
record_format arxiv
spellingShingle Consistent DAG selection for Bayesian causal discovery under general error distributions
Chaudhuri, Anamitra
Bhattacharya, Anirban
Ni, Yang
Statistics Theory
Methodology
Machine Learning
62H22, 62F15 (Primary) 62C10, 62E10 (Secondary)
We consider the problem of learning the underlying causal structure among a set of variables, which are assumed to follow a Bayesian network or, more specifically, a linear recursive structural equation model (SEM) with the associated errors being independent and allowed to be non-Gaussian. A Bayesian hierarchical model is proposed to identify the true data-generating directed acyclic graph (DAG) structure where the nodes and edges represent the variables and the direct causal effects, respectively. Moreover, incorporating the information of non-Gaussian errors, we characterize the distribution equivalence class of the true DAG, which specifies the best possible extent to which the DAG can be identified based on purely observational data. Furthermore, under the consideration that the errors are distributed as some scale mixture of Gaussian, where the mixing distribution is unspecified, and mild distributional assumptions, we establish that by employing a non-standard DAG prior, the posterior probability of the distribution equivalence class of the true DAG converges to unity as the sample size grows. This shows that the proposed method achieves the posterior DAG selection consistency, which is further illustrated with examples and simulation studies.
title Consistent DAG selection for Bayesian causal discovery under general error distributions
topic Statistics Theory
Methodology
Machine Learning
62H22, 62F15 (Primary) 62C10, 62E10 (Secondary)
url https://arxiv.org/abs/2508.00993