Propagation using Chain Event Graphs

TL;DR

This paper introduces a probability propagation algorithm for Chain Event Graphs (CEGs), leveraging their topology to efficiently handle asymmetric state spaces, offering an alternative to traditional Bayesian Network methods.

Contribution

The paper presents a novel propagation algorithm for CEGs that constructs a transporter CEG, analogous to junction trees in Bayesian Networks, improving efficiency in asymmetric problems.

Findings

The method is more efficient than BN algorithms in asymmetric contexts.

The algorithm uses factorization formulas tailored for CEG topology.

Transporter CEGs serve as a key structure for propagation.

Abstract

A Chain Event Graph (CEG) is a graphial model which designed to embody conditional independencies in problems whose state spaces are highly asymmetric and do not admit a natural product structure. In this paer we present a probability propagation algorithm which uses the topology of the CEG to build a transporter CEG. Intriungly,the transporter CEG is directly analogous to the triangulated Bayesian Network (BN) in the more conventional junction tree propagation algorithms used with BNs. The propagation method uses factorization formulae also analogous to (but different from) the ones using potentials on cliques and separators of the BN. It appears that the methods will be typically more efficient than the BN algorithms when applied to contexts where there is significant asymmetry present.