A modified simulation scheme for inference in Bayesian networks

In this paper we introduce an approximation method for uncertainty propagation which is based on a modification of the stratified simulation. The method uses a deterministic or perfect sample and calculates the number of times simulated instantiations are selected avoiding the repetition of identical instantiations which occurs in the standard stratified simulation method. A theoretical analysis to evaluate the performance of the method as compared with the stratified simulation scheme and to select the required step for the estimation of probabilities with a given error is presented. Some experimental studies compare the proposed with other simulation methods and show a much better performance not only in computer time but also in simulation errors.