SARIMA模型在济南市手足口病预警预测中的应用  

Application of SARIMA model in the prediction of hand, foot, and mouth disease

作　　者：常彩云[1] 许华茹[1] 赵梦娇[1] 徐淑慧[1] 成洪旗[1] 

机构地区：[1]山东省济南市疾病预防控制中心传染病与地方病防治所,250021

出　　处：《国际病毒学杂志》2017年第6期397-401,共5页International Journal of Virology

基　　金：济南市卫生局科技计划项目（2013-34）

摘　　要：目的 探讨季节性自回归移动平均（SARIMA）模型在手足口病月发病率预测中的应用,并利用该模型预测手足口病月发病趋势,为防控策略调整提供依据.方法 应用济南市2009-2016年手足口病月发病率资料建立SARIMA季节乘积模型,预测2017年1-7月发病率.结果 构建SARIMA（0,1,2）（0,l,1） 12模型可以用于济南市手足口病月发病率的拟合和预测,模型决定系数R2为0.830,均方根误差为6.355,标准化的BIC值为3.965.结论 建立的SARIMA模型能较好地拟合和预测手足口病的月发病数,为手足口病的防治提供参考依据.Objective In order to provide evidence for disease control and prevention,we explored the application of seasonal autoregressive integrated moving average （SARIMA） model for the incidence of hand,foot,and mouth disease （HFMD）,and predicted the monthly incidences of the disease.Methods The monthly incidences of HFMD from 2009 to 2016 were used to set up SARIMA prediction model.Data from January to July 2017 were used to validate the model and the prediction precision.Results SARIMA （0,1,2） （0,1,1） 12 model could be used to predict epidemiological tendency of HFMD in Jinan.The R2 of the model was 0.830,and the root mean square error was 6.355,and the standard BIC was 3.965.Conclusions The established SARIMA model had good fitting and prediction power for the monthly incidence of HFMD in Jinan,and can be used in the prevention and control of HFMD.

关 键 词：手足口病 SARIMA模型 预测 

分 类 号：R512.5[医药卫生—内科学]