The Bayesian network of being used forecasting

In this paper the method of constructing a Bayesian network is introduced and it is summarized what is the characteristic of Bayesian network and the content and process of learning Bayesian network,meanwhile the forecasting formula is introduced on discrete Bayesian network.There are three basic steps in learning Bayesian network,first is to determine the set and domain of variables;second is to determine the structure of network;third is to determine the local probabilistic distribution.Bayesian network is the graphical model of encoding the assertions of qualitative and quantitative relationships between variables and is a very useful tool for joint analysis and forecasting.