Independence of Causal Influence and Clique Tree Propagation

TL;DR

This paper investigates how independence of causal influence (ICI) can be exploited in clique tree propagation to improve the efficiency of Bayesian network inference, demonstrating significant empirical gains.

Contribution

It introduces a method to incorporate ICI into clique tree propagation, enhancing inference efficiency over previous techniques.

Findings

The new algorithm is significantly faster than previous methods.

Empirical results confirm improved computational efficiency.

ICI exploitation reduces complexity in Bayesian network inference.

Abstract

This paper explores the role of independence of causal influence (ICI) in Bayesian network inference. ICI allows one to factorize a conditional probability table into smaller pieces. We describe a method for exploiting the factorization in clique tree propagation (CTP) - the state-of-the-art exact inference algorithm for Bayesian networks. We also present empirical results showing that the resulting algorithm is significantly more efficient than the combination of CTP and previous techniques for exploiting ICI.