指数平滑灰色预测模型在HFRS发病率预测中的应用  

Application of exponential smoothing grey model in predicting the incidence of hemorrhagic fever with renal syndrome

作　　者：梁会营[1] 李雪莲[1] 郭军巧[2] 尹智华[1] 周宝森[1] 

机构地区：[1]中国医科大学公共卫生学院,辽宁沈阳110001 [2]辽宁省疾病预防控制中心,辽宁沈阳110001

出　　处：《中国热带医学》2009年第2期351-353,共3页China Tropical Medicine

基　　金：国家自然科学基金资助项目(30771860)

摘　　要：目的探讨肾综合征出血热(HFRS)发病率的预测方法,为合理调配肾综合征出血热防治的卫生资源提供依据。方法以1990～2001年辽宁省、丹东市和沈阳市HFRS的发病率为内样本建立指数平滑灰色预测模型I-GM(1,1),对2002年3个地区的HFRS发病进行预测,评价该模型的拟合和预测效果;并对3个地区未来几年的发病趋势进行预测。结果针对辽宁省HFRS发病率所建I-GM(1,1)模型的平均误差率(MER)和决定系数R2分别为14.73%,0.8573;针对丹东市HFRS发病率所建I-GM(1,1)模型的MER和R2分别为20.00%和0.8010;针对沈阳市HFRS发病率所建I-GM(1,1)模型的MER和R2分别为22.62%和0.8936。结论与GM(1,1)模型相比,指数平滑灰色预测模型I-GM(1,1)在预测疾病流行趋势方面是更加合理的选择;辽宁省HFRS发病率呈上升趋势,应采取有效的措施预防HFRS的发病。Objective To discuss the methods for predicating the incidence of hemorrhagic fever with renal syndrome （ HFRS） and provide scientific base for reasonable distribution of health resource of HFRS control. Methods I-GM（1,1）） model was established with the HFRS incidence of Liaoning Province,and the cities of Dandong and Shenyang from 1990 to 2001. The forecast of incidence of HFRS was made in the three areas in 2002,and the effects of fitting and forecasting were compared between two models. The trend of change of the incidence of HFRS in the coming several years was predicted. Results The MER of I-GM（1,1）model for Liaoning Province was 14.73%, and the R2 of the model was 0.8573. The MER of I-GM（1,1） model for Dandong was 20.00%, and the R2 of the model was 0.8010. The MER of I-GM（1,1）model for Shenyang was 22.62%, and the P,.2 of the model was 0.8936. Conclusion Comparing with the model of GM（1,1）, I-GM（1,1） model is a better choice. It also showed that the incidence of the HFRS revealed rising trend, thus preventive measures be taken for reducing the incidence of HFRS.

关 键 词：肾综合征出血热 GM(1 1)模型 指数平滑灰色预测模型 预测 

分 类 号：R-33[医药卫生]