Advances in Probabilistic Reasoning

TL;DR

This paper explores new methods for representing and reasoning with asymmetric independence in Bayesian networks, introducing faster inference, simplified network definitions, and a more expressive representation scheme.

Contribution

It presents a novel inference mechanism leveraging asymmetric independence, simplifies similarity network definitions, and generalizes representation schemes for Bayesian networks.

Findings

Inference speed improved by exploiting asymmetric independence

Simplified and extended similarity network definitions

A generalized scheme encoding more asymmetric independence assertions

Abstract

This paper discuses multiple Bayesian networks representation paradigms for encoding asymmetric independence assertions. We offer three contributions: (1) an inference mechanism that makes explicit use of asymmetric independence to speed up computations, (2) a simplified definition of similarity networks and extensions of their theory, and (3) a generalized representation scheme that encodes more types of asymmetric independence assertions than do similarity networks.