Graphical Representations for Algebraic Constraints of Linear Structural Equations Models

Thijs van Ommen; Mathias Drton

Graphical Representations for Algebraic Constraints of Linear Structural Equations Models

Sub Algorithmic Data Analysis

Research output: Chapter in Book/Report/Conference proceeding › Conference contribution › Academic › peer-review

Abstract

The observational characteristics of a linear structural equation model can be effectively described by polynomial constraints on the observed covariance matrix. However, these polynomials can be exponentially large, making them impractical for many purposes. In this paper, we present a graphical notation for many of these polynomial constraints. The expressive power of this notation is investigated both theoretically and empirically.

Original language	English
Title of host publication	Proceedings of The 11th International Conference on Probabilistic Graphical Models
Editors	Antonio Salmerón, Rafael Rumı́
Publisher	PMLR
Pages	409-420
Number of pages	12
Volume	186
Publication status	Published - 19 Sept 2022

Keywords

structural equation model
mixed graph
multivariate Gaussian
algebraic constraint

Access to Document

ommen22aFinal published version, 242 KB

https://proceedings.mlr.press/v186/ommen22a

Cite this

@inproceedings{9c14cca24f084539b83d06d2aabeb5a6,

title = "Graphical Representations for Algebraic Constraints of Linear Structural Equations Models",

abstract = "The observational characteristics of a linear structural equation model can be effectively described by polynomial constraints on the observed covariance matrix. However, these polynomials can be exponentially large, making them impractical for many purposes. In this paper, we present a graphical notation for many of these polynomial constraints. The expressive power of this notation is investigated both theoretically and empirically.",

keywords = "structural equation model, mixed graph, multivariate Gaussian, algebraic constraint",

author = "{van Ommen}, Thijs and Mathias Drton",

year = "2022",

month = sep,

day = "19",

language = "English",

volume = "186",

pages = "409--420",

editor = "Antonio Salmer{\'o}n and Rafael Rum{\'ı}",

booktitle = "Proceedings of The 11th International Conference on Probabilistic Graphical Models",

publisher = "PMLR",

}

Graphical Representations for Algebraic Constraints of Linear Structural Equations Models. / van Ommen, Thijs; Drton, Mathias.
Proceedings of The 11th International Conference on Probabilistic Graphical Models. ed. / Antonio Salmerón; Rafael Rumı́. Vol. 186 PMLR, 2022. p. 409-420.

Research output: Chapter in Book/Report/Conference proceeding › Conference contribution › Academic › peer-review

TY - GEN

T1 - Graphical Representations for Algebraic Constraints of Linear Structural Equations Models

AU - van Ommen, Thijs

AU - Drton, Mathias

PY - 2022/9/19

Y1 - 2022/9/19

N2 - The observational characteristics of a linear structural equation model can be effectively described by polynomial constraints on the observed covariance matrix. However, these polynomials can be exponentially large, making them impractical for many purposes. In this paper, we present a graphical notation for many of these polynomial constraints. The expressive power of this notation is investigated both theoretically and empirically.

AB - The observational characteristics of a linear structural equation model can be effectively described by polynomial constraints on the observed covariance matrix. However, these polynomials can be exponentially large, making them impractical for many purposes. In this paper, we present a graphical notation for many of these polynomial constraints. The expressive power of this notation is investigated both theoretically and empirically.

KW - structural equation model

KW - mixed graph

KW - multivariate Gaussian

KW - algebraic constraint

M3 - Conference contribution

VL - 186

SP - 409

EP - 420

BT - Proceedings of The 11th International Conference on Probabilistic Graphical Models

A2 - Salmerón, Antonio

A2 - Rumı́, Rafael

PB - PMLR

ER -

Graphical Representations for Algebraic Constraints of Linear Structural Equations Models

Abstract

Keywords

Access to Document

Fingerprint

Cite this