应用时间序列模型预测宜昌市甲型病毒性肝炎发病率

Study on the appliance of Time Series Model for the prediction of the incidence of viral hepatitis type A

目的探讨应用ARIMA、季节趋势等时间系列模型预测甲型病毒性肝炎发病率。方法基于2005—2015年宜昌市逐月甲型病毒性肝炎发病率建立两种模型,对2016年甲型病毒性肝炎的发病率进行预测;采用平均绝对百分误差(MAPE)对预测模型进行拟合效果评价。结果两种时间序列与季节趋势模型能较好的模拟甲肝发病,但时间序列模型优于季节性趋势。应用最优模型对2016年甲型病毒性肝炎的发病率进行预测。结论两种时间序列与季节趋势模型能较好的模拟宜昌市甲型病毒性肝炎发病在时间序列的变化趋势,但ARIMA(0,0,1)(0,1,1)12模型能较好的模拟甲型病毒性肝炎发病在时间序列的变化趋势,为制定科学的防控措施和策略提供依据。

Objective To explore the application of two kinds of Time Series Models and Seasonal Trend Model in predicting the incidence of viral hepatitis type A.Methods Two kinds of models were built based on the monthly incidence of viral hepatitis type A from 2005 to 2015 in Yichang City to predict the incidence in 2016,MAPE method was used to evaluate the fitting effect between the actual value and the predicted value.Results Two kinds of Time Series Models,as well as Seasonal Trend Model were be able to well simulate the change trend of time series of viral hepatitis type A,Time Series Models were better than Seasonal Trend Model.The best model would be used to predict the incidence of viral hepatitis A.Conclusion Two kinds of Time Series Models,as well as Seasonal Trend Model were be able to well simulate the change trend of time series of viral hepatitis type A.However,the ARIMA（0,0,1）（0,1,1） 12 models could better simulate the change trend of time series of viral hepatitis type A incidence,and provide evidence for establishing control strategy and measures.