Hyperplane Representations of Interventional Characteristic Imset Polytopes

Fuente: arXiv
Salvato in:
Dettagli Bibliografici
Autori principali: Hollering, Benjamin, Johnson, Joseph, Solus, Liam
Natura: Preprint
Pubblicazione: 2024
Soggetti:
Accesso online:
Tags: Aggiungi Tag
Nessun Tag, puoi essere il primo ad aggiungerne!!
_version_ 1866909183982436352
author Hollering, Benjamin
Johnson, Joseph
Solus, Liam
author_facet Hollering, Benjamin
Johnson, Joseph
Solus, Liam
contents Characteristic imsets are 0/1-vectors representing directed acyclic graphs whose edges represent direct cause-effect relations between jointly distributed random variables. A characteristic imset (CIM) polytope is the convex hull of a collection of characteristic imsets. CIM polytopes arise as feasible regions of a linear programming approach to the problem of causal disovery, which aims to infer a cause-effect structure from data. Linear optimization methods typically require a hyperplane representation of the feasible region, which has proven difficult to compute for CIM polytopes despite continued efforts. We solve this problem for CIM polytopes that are the convex hull of imsets associated to DAGs whose underlying graph of adjacencies is a tree. Our methods use the theory of toric fiber products as well as the novel notion of interventional CIM polytopes. Our solution is obtained as a corollary of a more general result for interventional CIM polytopes. The identified hyperplanes are applied to yield a linear optimization-based causal discovery algorithm for learning polytree causal networks from a combination of observational and interventional data.
format Preprint
id arxiv_https___arxiv_org_abs_2404_18500
institution arXiv
publishDate 2024
record_format arxiv
spellingShingle Hyperplane Representations of Interventional Characteristic Imset Polytopes
Hollering, Benjamin
Johnson, Joseph
Solus, Liam
Combinatorics
Statistics Theory
62E10, 62H22, 62D20, 62R01, 13P25, 13P10
Characteristic imsets are 0/1-vectors representing directed acyclic graphs whose edges represent direct cause-effect relations between jointly distributed random variables. A characteristic imset (CIM) polytope is the convex hull of a collection of characteristic imsets. CIM polytopes arise as feasible regions of a linear programming approach to the problem of causal disovery, which aims to infer a cause-effect structure from data. Linear optimization methods typically require a hyperplane representation of the feasible region, which has proven difficult to compute for CIM polytopes despite continued efforts. We solve this problem for CIM polytopes that are the convex hull of imsets associated to DAGs whose underlying graph of adjacencies is a tree. Our methods use the theory of toric fiber products as well as the novel notion of interventional CIM polytopes. Our solution is obtained as a corollary of a more general result for interventional CIM polytopes. The identified hyperplanes are applied to yield a linear optimization-based causal discovery algorithm for learning polytree causal networks from a combination of observational and interventional data.
title Hyperplane Representations of Interventional Characteristic Imset Polytopes
topic Combinatorics
Statistics Theory
62E10, 62H22, 62D20, 62R01, 13P25, 13P10
url https://arxiv.org/abs/2404.18500