Inference Algorithms for Similarity Networks

TL;DR

This paper introduces efficient inference algorithms for two types of similarity networks based on relevance, with one algorithm applicable without restrictions, enhancing probabilistic reasoning in such networks.

Contribution

It presents novel inference algorithms for two types of similarity networks, including a more general algorithm that requires no assumptions about event probabilities.

Findings

Efficient inference algorithms for both types of similarity networks.

A more general inference algorithm for type 1 networks without restrictions.

Algorithms assume all events have nonzero probability, with an alternative less efficient method.

Abstract

We examine two types of similarity networks each based on a distinct notion of relevance. For both types of similarity networks we present an efficient inference algorithm that works under the assumption that every event has a nonzero probability of occurrence. Another inference algorithm is developed for type 1 similarity networks that works under no restriction, albeit less efficiently.