Approximate Inference Algorithms for Hybrid Bayesian Networks with Discrete Constraints

TL;DR

This paper introduces two approximate inference algorithms for Hybrid Mixed Networks that incorporate discrete constraints, combining belief propagation, importance sampling, and constraint propagation to improve reasoning in complex models.

Contribution

The paper presents novel approximate inference algorithms specifically designed for Hybrid Mixed Networks with discrete constraints, integrating multiple algorithmic principles.

Findings

Algorithms perform well on randomly generated HMNs

Effective integration of belief propagation and importance sampling

Improved reasoning in hybrid models with constraints

Abstract

In this paper, we consider Hybrid Mixed Networks (HMN) which are Hybrid Bayesian Networks that allow discrete deterministic information to be modeled explicitly in the form of constraints. We present two approximate inference algorithms for HMNs that integrate and adjust well known algorithmic principles such as Generalized Belief Propagation, Rao-Blackwellised Importance Sampling and Constraint Propagation to address the complexity of modeling and reasoning in HMNs. We demonstrate the performance of our approximate inference algorithms on randomly generated HMNs.