ARIMA模型预测上海市闸北区手足口病发病趋势  

Application of ARIMA Model in Predicting Incidence Trend of Hand-foot-mouth Disease in Zhabei District,Shanghai

作　　者：蔡晓虹[1] 万秋萍[1] 吴益生[1] 熊建箐 徐敏钢 李恩国[1] 

机构地区：[1]上海市闸北区疾病预防控制中心,上海200072

出　　处：《实用预防医学》2012年第3期381-384,共4页Practical Preventive Medicine

摘　　要：目的应用季节性时间序列ARIMA模型建立手足口病发病趋势预测,为预警、早期防控手足口病流行提供依据。方法应用SPSS13.0对2002年4月-2011年3月8年的手足口病逐月发病率建立ARIMA模型。结果通过对参数和模型的拟合优度检验以及残差白噪声序列的检验,最终确定模型为ARIMA(1,0,0)(2,1,0)12,其中AIC=235.855,BIC=245.942,LB统计量检验残差序列为白噪声序列。结论模型能够有效地预测手足口病发病趋势,对预警预测产生积极的指导作用。Objective To establish the seasonal time series ARIMA model for forecasting the epidemic trends of hand-foot-mouth disease（HFMD）,and to provide the evidence for early warning,prevention and control of HFMD at an early stage. Methods The ARIMA model was established based on the monthly incidence rates of HFMD in Zhabei District,Shanghai from April 2002 to March 2011 by SPSS 13.0 software. Results Through the test of parameters and goodness of fit as well as whitenoise residuals,we finalized the model ARIMA（1,0,0）（2,1,0）12,of which AIC（Akaike Information Criterion）=235.855,BIC（Bayesian Information Criterion）=245.942. Conclusions The model can predict the incidence trend of hand-foot-mouth disease,and moreover,it plays a positive role in guiding for early warning and forecast.

关 键 词：时间序列分析 ARIMA模型 手足口病 预测 

分 类 号：R512.5[医药卫生—内科学]