Web marketing can improve the management level of power-grid enterprises. With the increasing of market-oriented process, it is necessary to classify the customer efficiently to provide special electricity services. The predictive-modeling technologies can carry out classification for customer well. Thus, the four typical predictive-modeling technologies--regression analysis, Bayesian networks, decision tree, and neural networks--are introduced. In addition, the instances in web power marketing are used for illustrating these methods. The application of one or more predictive-modeling technologies can classify well to improve the marketing efficient of power-grid enterprises.
Neural Networks and Their Application to Power Engineering