Factor Analysis Of Bayesian Network Method

As a probability model, Bayesian network is an effective tool for the uncertainty of the simulation and reasoning, which make use of conditional independent relationship among a set of random variables to reduce the number of the parameters that are needed for the joint probability. With the ability of changing with the new information, Bayesian network can be integrated by data and prior knowledge of experts, and then identify potential links and correlation among variables. The purpose of finding Bayesian network structure is to seek the structure that matches with data and prior knowledge well. Bayesian network is practical in a wide range of applications. In this paper, Bayesian network method is compared with the traditional factor analysis,and example shows the result of Bayesian network method can be better than or equivalent to the traditional factor analysis. On the other hand, as a kind of nonlinear factor analysis, Bayesian network method has a much wider application. And at last we have a deep discussion for the classification.