SARIMA模型在广东省痢疾发病预测中的应用  

Applications of SARIMA Model on Forecast Incidence of Bacillary Dysentery in Guangdong

作　　者：许可欣[1] 王永波[1] 钱俊[1] 

机构地区：[1]南方医科大学生物医学工程学院,广州510515

出　　处：《数理医药学杂志》2015年第11期1584-1587,共4页Journal of Mathematical Medicine

基　　金：广东省科技计划项目(2013B060500046);广东省大学生创新训练项目(201412121062)

摘　　要：目的：探讨SARIMA模型分析预测广东省痢疾发病数的可行性和适用性,为监控及早期防控提供依据.方法：对2009年1月-2014年6月广东省痢疾月发病例数数据资料建立SARIMA模型,并以2014年7月-2014年12月发病例数验证模型.结果：通过对参数估计和模型的拟合优度检验以及白噪声检验,建立模型SARIMA（1,1,0）（0,1,1）12,其预测值与实际值变动趋势相同,平均相对误差为6.78%.结论：SARIMA（1,1,0）（0,1,1）12模型可用于模拟广东省痢疾发病的变化趋势并进行短期预测.Objective:To explore the feasibility and application of autoregressive integrated moving aver-age(ARIMA)model to forecast the incidence of bacillary dysentery in Guangdong.Methods:SARIMA model was fitted with data of monthly incidence from January 2009 to June 2014 for dysentery and the monthly inci-dence of July to December in 2014 for dysentery was predicted and evaluated.Results:Through the goodness of fit test and white noise test among the residual for the model,SARIMA(1,1,0)(0,1,1)12 was established. The forecast monthly incidence of dysentery from January to December in 2014 consisted with the actual one, the relative error was 6.78%.Conclusion:SARIMA(1,1,0)(0,1,1)12 can be used to fit trends of incidence for dysentery in Guangdong province and forecast within short period.

关 键 词：SARIMA模型 痢疾 预测 SAS 

分 类 号：R311[医药卫生—基础医学]