Designing neural networks that process mean values of random variables

TL;DR

This paper introduces neural networks inspired by Bayesian networks that compute mean values of random variables without training, effectively pooling evidence and handling contradictions, thus bridging probabilistic models and neural computation.

Contribution

It presents a novel class of neural networks derived from Bayesian models that require no training and can perform probabilistic inference tasks.

Findings

Networks accurately compute mean values of variables.

They can pool multiple evidence sources.

They handle inconsistent evidence effectively.

Abstract

We introduce a class of neural networks derived from probabilistic models in the form of Bayesian networks. By imposing additional assumptions about the nature of the probabilistic models represented in the networks, we derive neural networks with standard dynamics that require no training to determine the synaptic weights, that perform accurate calculation of the mean values of the random variables, that can pool multiple sources of evidence, and that deal cleanly and consistently with inconsistent or contradictory evidence. The presented neural networks capture many properties of Bayesian networks, providing distributed versions of probabilistic models.