论文部分内容阅读
目的研究基于GRNN的组合预测模型拟合传染病发病率的优越性和不足。方法以浙中某市1998—2008年的肺结核发病率为研究资料,分别构建了灰色模型和ARIMA模型,以这两种模型为基础构建了基于GRNN的组合预测模型。结果残差修正GM(1,1)模型、ARIMA(1,0,1)*(1,1,0)12模型、基于GRNN的组合预测模型的MSE,MAE,MAPE和MER分别为37.451,5.692,53.69%,48.51%;18.509,3.761,35.13%,32.05%;9.961,2.571,25.6%,21.9%。结论基于GRNN的组合预测模型的预测精度优于两种单项模型。
Objective To study the advantages and disadvantages of the combined forecasting model based on GRNN to fit the incidence of infectious diseases. Methods Based on the prevalence of tuberculosis in a city of Zhejiang Province from 1998 to 2008 as the research data, the gray model and the ARIMA model were respectively constructed. Based on these two models, a combined forecasting model based on GRNN was constructed. Results The residuals modified GM (1,1) model, ARIMA (1,0,1) * (1,1,0) 12 model, combined forecasting model based on GRNN, MSE, MAE, MAPE and MER were 37.451,5.692 , 53.69%, 48.51%; 18.509,3.761,35.13%, 32.05%; 9.961, 2.571, 25.6%, 21.9% respectively. Conclusion The prediction accuracy of the combined forecasting model based on GRNN is better than that of two single models.