Forecast Of The Regional EC Development Through An ANN Model With A Feedback Controller
Free (open access)
G. Jianquan, Fankun, T. Bingyong, B. Shi & Y. Jianzheng
This paper is to have a deep understanding of the way to forecast the economic development with the help of an Artificial Neural Network (ANN), putting forward a brand-new ANN forecast model, that is, the Back Propagation Networking Learning Algorithm (BP Networking Algorithm) with a feedback controller. The model has been used to overcome the deficiencies of the traditional BP Algorithm, as it is more accurate for forecasting, less dependable on initial data, and easier to select the needed number of hidden layers and hidden-layer neurons. In order to measure regional electronic commerce development we have set an evaluation system, which seems to be comparatively perfect and manipulative. With the model and the system, we carried out a regional EC forecast in Huai’nan, a medium-sized city in Anhui Province, China. The result of the case study has indicated that the model has an ideal extension, the number of its hidden-layer neurons can easily be decided, and we are to have a long-term forecast of the development without much initial data. With this model in hand, it is possible to cope with the problems of sparse, dispersed and hard-to-forecast statistical information in the development of electronic commerce. Keywords: feedback controller, BP model, EC development, forecast.
feedback controller, BP model, EC development, forecast.