应用ARIMA模型对甲型病毒性肝炎疫情预测的研究

目的探讨时间序列ARIMA模型在浙江省甲型病毒性肝炎疫情预警应用中的可行性。方法利用EViews5.0、SPSS13.0等统计软件对历年疫情资料汇编,《国家疾病报告管理信息系统》报告的浙江省甲肝按月发病数建立ARIMA模型。结果对甲肝发病数序列建立季节模型为ARIMA(1,0,0)(0,1,1)12,对2009年甲肝按月发病疫情拟合较好,能够及时、准确对疫情进行跟踪。结论 ARIMA模型能够较好应用于短期甲肝预测、预警,为疫情防控提供科学依据。

Application of ARIMA Model on Prediction of the Incidence of Hepatitis A

Objective This study was conducted to explore the feasibility of time-series ARIMA model on prediction of hepatitis A attack.Methods EViews5 and SPSS 13.0 statistical software were used to fit ARIMA mode1 by monthly incidence of hepatitis a from 1990 January to 2008 December in Zhejiang Province.Results The season model ARIMA（1,0,0）（0,1,1） was fit for the prediction of monthly incidence of hepatitis A in 2009,and can timely and accurately trace the hepatitis A attack. Conclusion ARIMA model can be applied to make a short term prediction of the hepatitis A attack,providing a scientific evidence for its prevention and control.