基于RBF神经网络的总产值预测模型研究

Research on Predicting Model of Gross Production Value Using RBF Neural Network

介绍了径向基函数(RBF)神经网络的工作原理和训练算法．根据生产总值与投资分配之间存在的映射关系,应用RBF神经网络建立了投资预测模型,并进行了仿真试验．与BP模型相比,该模型在预测精度和收敛速度方面具有显著的优点．结果表明,用RBF神经网络进行投资预测得到了十分满意的结果．

The work theory and training algorithm of RBF Neural Network were presented in this paper. A predicting model of investment was established on the mapped relationship between gross production value and distribution of investment using RBF neural Network and a simulation was done. Compared with BP model, it had apparent advantages in predicting precision and convergence rate. The predicting result indicated that it was well -content to predict using RBF Neural Network.