基于灰色G(1,1)和小波神经网络的预测模型及应用

将灰色系统、小波分析和三层BP神经网络各自优点集于一身建立了基于灰色G(1,1)和小波神经网络的预测模型,大幅度提高了模型的预测精度和可靠性。选用我国自1994年至2006年狂犬病发病率统计数据,用灰色GM(1,1)模型对历年的疾病发病人数进行建模,将拟合值做小波神经网络的输入进行二次拟合和预测。实验结果及仿真验证表明,本文模型预测效果远优于单一的灰色模型预测。

Based on gray G(1,1) and wavelet neural network model predictions and applications

Summary of the gray system,wavelet analysis and three-layer BP neural network has established based on their respective merits rolled into one gray G(1,1) and wavelet neural network prediction model significantly improved model prediction accuracy and reliability.Used in China since 1994 to 2006 the incidence of rabies statistics,gray GM(1,1) model over the years to model the incidence of the disease,the fitted value of the input wavelet neural network to do a second fitting and prediction.Experimental resu...