Algebraic Constraints for Linear Acyclic Causal Models

Fuente: arXiv
Saved in:
Bibliographic Details
Main Authors: Gigliotti, Cole, Robeva, Elina
Format: Preprint
Published: 2025
Subjects:
Online Access:
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1866915368238317568
author Gigliotti, Cole
Robeva, Elina
author_facet Gigliotti, Cole
Robeva, Elina
contents In this paper we study the space of second- and third-order moment tensors of random vectors which satisfy a Linear Non-Gaussian Acyclic Model (LiNGAM). In such a causal model each entry $X_i$ of the random vector $X$ corresponds to a vertex $i$ of a directed acyclic graph $G$ and can be expressed as a linear combination of its direct causes $\{X_j: j\to i\}$ and random noise. For any directed acyclic graph $G$, we show that a random vector $X$ arises from a LiNGAM with graph $G$ if and only if certain easy-to-construct matrices, whose entries are second- and third-order moments of $X$, drop rank. This determinantal characterization extends previous results proven for polytrees and generalizes the well-known local Markov property for Gaussian models.
format Preprint
id arxiv_https___arxiv_org_abs_2505_00215
institution arXiv
publishDate 2025
record_format arxiv
spellingShingle Algebraic Constraints for Linear Acyclic Causal Models
Gigliotti, Cole
Robeva, Elina
Statistics Theory
62R01, 62H22
In this paper we study the space of second- and third-order moment tensors of random vectors which satisfy a Linear Non-Gaussian Acyclic Model (LiNGAM). In such a causal model each entry $X_i$ of the random vector $X$ corresponds to a vertex $i$ of a directed acyclic graph $G$ and can be expressed as a linear combination of its direct causes $\{X_j: j\to i\}$ and random noise. For any directed acyclic graph $G$, we show that a random vector $X$ arises from a LiNGAM with graph $G$ if and only if certain easy-to-construct matrices, whose entries are second- and third-order moments of $X$, drop rank. This determinantal characterization extends previous results proven for polytrees and generalizes the well-known local Markov property for Gaussian models.
title Algebraic Constraints for Linear Acyclic Causal Models
topic Statistics Theory
62R01, 62H22
url https://arxiv.org/abs/2505.00215