摘要
目的探讨变权组合模型在我国手足口病月发病率预测中的应用,并与ARIMA模型和残差自回归模型的预测效果进行比较。方法收集2008年1月至2014年12月我国手足口病月发病率资料,用SPSS13.0和Eviews8.0拟合三种模型,并用2014年7-12月的数据比较三种模型的拟合和预测效果。结果 ARIMA模型、残差自回归模型和变权组合模型拟合及预测的MRD,MSE,RMSE和MAE分别为14.006,4.689,2.165,0.147和13.565,4.416,2.101,0.133;16.793,7.247,2.692,0.171和16.206,6.639,2.577,0.164;8.447,1.843,1.358,0.092和8.409,1.833,1.354,0.082。结论变权组合模型拟合及预测效果优于ARIMA模型和残差自回归模型。
Objective To explore the application of ARIMA model and auto-regressive model and weight changeable model in prediction on incidence of HFM D in China and compare the predictive effect among them. Methods The data of monthly incidence of HFM D from January 2008 to December 2014 in China w as collected and Spss13. 0 and Eview s8. 0 w ere used to fit three models. Afterw ards,the data from July to December in 2014 w as used to evaluate the predictive effect. Results The M RD,M SE,RM SE and M AE fitted and predicted by ARIM A model and Auto-Regressive model and w eight changeable model w ere 4. 006,4. 689,2. 165,0. 147 and 13. 565,4. 416,2. 101,0. 133; 16. 793,7. 247,2. 692,0. 171 and 16. 206,6. 639,2. 577,0. 164; 8. 447,1. 843,1. 358,0. 092 and 8. 409,1. 833,1. 354,0. 082,respectively. Conclusions According to the model fitness and prediction accuracy,the w eight changeable model is superior to the ARIM A model and Auto-Regressive model w ith a good practical value.
出处
《中国卫生统计》
CSCD
北大核心
2016年第3期412-415,共4页
Chinese Journal of Health Statistics
基金
河北省卫生厅医学科学研究重点课题计划(20130055)