The Bayesian Network Method for Scenario Prediction

Bayesian network is a powerful tool for scenario analysis and prediction. At present, future scenario is structured by combining the results of forecasting every variable. The dependency relationship between variables is neglected so that the result of scenario prediction often is unreliable. In this paper, a Bayesian network is built by combining subjective expert knowledge and objective data and a scenario is forecasted by joint probability calculation based on Bayesian network. Because the dependency relationship between variables can be effectively used, forecasting result will be more believable and prediction and analysis on scenario can also combined.