ARIMA模型在乙型肝炎发病率预测中的应用

目的探讨ARIMA模型预测常州市乙型肝炎发病率的可行性。方法应用SPSS 18.0软件对常州市2002—2010年乙型肝炎月发病率进行ARIMA模型拟合,对2011年乙型肝炎发病率进行预测验证,并对2012年月发病率进行预测。结果最佳拟合模型为ARIMA(2,1,0)(1,1,0)12;2011年乙型肝炎发病率预测结果和实际值平均相对误差为3.55%。结论 ARIMA模型可用于常州市乙型肝炎发病率预测,其短期预测精度较高。

The Application of ARIMA Model on Prediction of the Incidence of Hepatitis B

Objective To discuss the feasibility of prediction of the incidence of Hepatitis B in Changzhou by ARIMA Model.Methods Incidence of Hepatitis B was collected monthly from 2002 to 2010 to fit a ARIMA model by using SPSS 18.0,The model was predicted and verified by the incidence of Hepatitis B in 2011 and was used to the predict the month incidence in 2012.Results The best fitting model was ARIMA(2,1,0)(1,1,0)12 and the mean relative error between the predicted value and actual value of the incidence of Hepatitis B in 2011 was 3.55%.Conclusion The ARIMA model could be used to predict the incidence of Hepatitis B in Changzhou with relatively high short-term prediction accuracy.