The Application Of Neural Networks In The Chinese Telecom Value-added Services

In recent years, the communication industry development is swift and violent, compares with the high speed growth user scale and the telephone conversation total quantity, the income grows quite slowly, appeared "The increase does not additionally receive" the phenomenon, the customer ARPU value reduced continually. Specially XX area by 2006 years year's end CHINA MOBILE and CHINA UNITED moves a series of promotion move, has seized very many customer resources from the Telecommunication. The Telecommunication customer number obvious reduces large scale. But actually maintains exuberant growth tendency in the last few years with it correspondence Telecommunication Value-Added Services, the income increases year by year. Therefore carries on the thorough analysis to the Value-Added Service, enhances it to occupy the ratio in the income, sped up its income rate of increment then to become us to solve the question the most important Operation business should develop the service new domain positively, carries on the service market forecast, guarantees the enterprise to receive and the profit growth, enhance the enterprise market competition ability and sustainable development ability.The paper mainly revolves the XX Telecommunication Value-Added service classification and the forecast model research carries on launches. Through SOM constructs PAS user's classified model discovers each kind of user community characteristic, again to has the representative some kind of user to carry on the correlation analysis to conduct the further research to this kind of user's characteristic, and discovers the marketing countermeasure. Finally introduced how makes the improvement using the genetic algorithm to the BP neural network, and carries on the forecast modeling using the improvement BP neural network to some Value-Added service income, provides the correlation basis for the Telecommunication effective formulation decision-making.The paper has mainly included the data preparation, the model establishment, model implementation and the appraisal process. By the XX Telecommunication Value-Added service took the object of study, by the XX Telecommunication operation actual transport business data took the research data, studies the tool by MATLAB and the SPSS software achievement, takes the master line by the neural network in the Value-Added service utilization, carries on a series of analyses to the Value-Added service, discovers supports the decision-making to the Value-Added service marketing useful information.