Global Conditioning for Probabilistic Inference in Belief Networks

TL;DR

This paper introduces global conditioning, a new generalization of loop-cutset conditioning for probabilistic inference in belief networks, enabling parallel processing and memory-time tradeoffs within a unified framework.

Contribution

It presents a unifying framework that encompasses existing methods and introduces global conditioning as a flexible, parallelizable approach for inference in belief networks.

Findings

Global conditioning generalizes loop-cutset conditioning.

The approach enables parallel processing and memory-time tradeoffs.

A hybrid method combining existing techniques is also unified.

Abstract

In this paper we propose a new approach to probabilistic inference on belief networks, global conditioning, which is a simple generalization of Pearl's (1986b) method of loopcutset conditioning. We show that global conditioning, as well as loop-cutset conditioning, can be thought of as a special case of the method of Lauritzen and Spiegelhalter (1988) as refined by Jensen et al (199Oa; 1990b). Nonetheless, this approach provides new opportunities for parallel processing and, in the case of sequential processing, a tradeoff of time for memory. We also show how a hybrid method (Suermondt and others 1990) combining loop-cutset conditioning with Jensen's method can be viewed within our framework. By exploring the relationships between these methods, we develop a unifying framework in which the advantages of each approach can be combined successfully.