Bayesian Networks Modeling Methods for Complex Problems
MA Xi-jun · Jisuanji fangzhen · 2006
Construction of a Bayesian network for a complex problem is a system engineering. Certain programmers must be followed. The process of Bayesian networks modeling for complex problems was built. Three phases are included: problem analyzing, model designing and model testing. The jobs which should be done in every phases were illustrated. And in the phase of model designing, a framework combining expert knowledge with data to construct a network was represented. At last, some simplified principles for modeling were suggested.