ARIMA模型在流行性腮腺炎疫情预测中的应用

Application of ARIMA model in predict outbreaks of mumps

目的建立乘积季节自回归移动平均（ARIMA）模型,观察其对河南省流行性腮腺炎疫情预测的可行性。方法利用河南省2004~2013年分月的流行性腮腺炎疫情监测资料建立乘积季节ARIMA模型,利用2014年1~12月的流行性腮腺炎疫情资料评价该模型的预测效能。结果河南省2004~2013年流行性腮腺炎发病呈现明显的季节效应,且发病数在2006年后呈现逐年增多的趋势;模型ARIMA（1,0,2）（0,1,1）12能较好地拟合既往的流行性腮腺炎报告病例数,且对2014年1~12月按月报告的流行性腮腺炎病例数的预测值与实际值基本吻合。结论 ARIMA模型能较好地模拟、预测河南省流行性腮腺炎的发病情况。

Objective To establish a multiplicative seasonal autoregressive integrated moving average （ARIMA） model, and observe the feasibility of prediction of mumps epidemic in Henan Province. Methods Mumps surveillance data in Henan Province from 2004 to 2013 Year-month was taken to establish multiplicative seasonal ARIMA model, mumps outbreak data from January 2014 to December 2014 was taken to evaluate the predictive power of the model. Results The incidence of mumps in Henan from 2004 to 2013 were showed obvious seasonal effects and the number of incidence in 2006 after showing increased year by year; model ARIMA （1, 0, 2） （0, 1, 1） 12 was indentical with the past number of reported cases of mumps, the predicted and actual values of January to December 2014 monthly report on the number of mumps basically. Conclusion ARIMA model can simulate, predict the incidence of mumps in He＇nan Province.