Library Header Image
LSE Research Online LSE Library Services

Lattice conditional independence models and Hibi ideals

Caines, Peter, Mohammadi, Fatemeh, Sáenz-de-Cabezón, Eduardo and Wynn, Henry ORCID: 0000-0002-6448-1080 (2022) Lattice conditional independence models and Hibi ideals. Transactions of the London Mathematical Society, 9 (1). 1 - 19. ISSN 2052-4986

[img] Text (Wynn_lattice-conditional-independence-models--published) - Published Version
Available under License Creative Commons Attribution Non-commercial No Derivatives.

Download (271kB)

Identification Number: 10.1112/tlm3.12041


Lattice conditional independence models [Andersson and Perlman, Lattice models for conditional independence in a multivariate normal distribution, Ann. Statist. 21 (1993), 1318–1358] are a class of models developed first for the Gaussian case in which a distributive lattice classifies all the conditional independence statements. The main result is that these models can equivalently be described via a transitive directed acyclic graph (TDAG) in which, as is normal for causal models, the conditional independence is in terms of conditioning on ancestors in the graph. We demonstrate that a parallel stream of research in algebra, the theory of Hibi ideals, not only maps directly to the lattice conditional independence models but gives a vehicle to generalise the theory from the linear Gaussian case. Given a distributive lattice (i) each conditional independence statement is associated with a Hibi relation defined on the lattice, (ii) the directed graph is given by chains in the lattice which correspond to chains of conditional independence, (iii) the elimination ideal of product terms in the chains gives the Hibi ideal and (iv) the TDAG can be recovered from a special bipartite graph constructed via the Alexander dual of the Hibi ideal. It is briefly demonstrated that there are natural applications to statistical log-linear models, time series and Shannon information flow.

Item Type: Article
Official URL:
Additional Information: © 2022 The Author(s).
Divisions: Statistics
Subjects: H Social Sciences > HA Statistics
Date Deposited: 03 Jan 2023 16:48
Last Modified: 09 Jan 2023 09:45

Actions (login required)

View Item View Item


Downloads per month over past year

View more statistics