Improvement of the junction tree algorithm for exact inference in dynamic Bayesian networks
Zhao Jian-zhong · Journal of Hefei University of Technology · 2006
In solving the dynamic uncertainty problem based on dynamic Bayesian networks(DBNs),the efficiency of processing is decided by the inference algorithm.In the paper,a 3/2 slice junction tree algorithm is presented in which the elimination order is not limited and constitution of the junction tree is carried out only one time.Comparative analysis shows that it is simpler and has a lower complexity.