Inter-causal Independence and Heterogeneous Factorization

TL;DR

This paper introduces the concept of inter-causal independence to enhance factorization in Bayesian networks, leading to more efficient inference algorithms by leveraging both conditional and inter-causal independence.

Contribution

It provides a constructive definition of inter-causal independence and develops an inference algorithm that exploits this alongside conditional independence.

Findings

The proposed algorithm reduces inference complexity in Bayesian networks.

Inter-causal independence enables further factorization of conditional probabilities.

The method improves computational efficiency in probabilistic inference.

Abstract

It is well known that conditional independence can be used to factorize a joint probability into a multiplication of conditional probabilities. This paper proposes a constructive definition of inter-causal independence, which can be used to further factorize a conditional probability. An inference algorithm is developed, which makes use of both conditional independence and inter-causal independence to reduce inference complexity in Bayesian networks.