Importance Sampling for General Hybrid Bayesian Networks
Changhe Yuan, Marek J. Drużdżel · 2007
Some real problems are more naturally mod-eled by hybrid Bayesian networks that consist of mixtures of continuous and discrete variables with their interactions described by equations and continuous probability distributions. How-ever, inference in such general hybrid models is hard. Therefore, existing approaches either only deal with special instances, such as Conditional Linear Gaussians (CLGs), or approximate a gen-eral model with a restricted version and then per-form inference on the simpler model. However, results thus obtained highly depend on the qual-ity of the approximations. This paper describes an importance sampling-based algorithm that di-rectly deals with hybrid Bayesian networks con-structed in the most general settings and guar-antees to converge to the correct answers given enough time. 1