Electric power enterprise has accumulated much data for many years, and needs to give reliable bases for sellers who make decisions through knowledge discovering. Bayesian method has been one of the most excellent focuses in knowledge discovering domain through its uncertain knowledge expression form, abundant probability expression ability and study characteristic with transcendent knowledge. Introducing Bayesian method to electric power marketing decision-making can get many actual effects that other methods can not match. This paper gives focus on the Bayesian network's application in data mining. It constructs Bayesian network which is suitable to electric power marketing decision-making combining with the problem that sellers concern for. These nets are used for evaluating customer's value, monitoring the disorder of utilizing electric and other parts, they perfect themselves through study algorithm. The experiments with actual data show that Bayesian network is practical and reliable in electric power marketing decision-making. |