ARIMA乘积季节模型在食源性疾病月发病数预测中的应用  

Application of seasonal ARIMA model in predicting the monthly incidence of foodborne diseases

作　　者：张雪佩 周林[3] 刘敏 滕艾颖 李艳花 马伟[2] ZHANG Xuepei;ZHOU Lin;LIU Min;TENG Aiying;LI Yanhua;MA Wei(Provincial Hospital Affiliated to Shandong First Medical University,Jinan,Shandong 250012,China;School of Public Health,Shandong University,Jinan,Shandong 250012;China 3.Jinan Center for Disease Control and Prevention,Jinan,Shandong 250021,China)

机构地区：[1]山东第一医科大学附属省立医院,山东济南250021 [2]山东大学公共卫生学院,山东济南250012 [3]山东省济南市疾病预防控制中心,山东济南250021

出　　处：《公共卫生与预防医学》2024年第5期6-9,共4页Journal of Public Health and Preventive Medicine

摘　　要：目的探讨济南市地区食源性疾病发病趋势特征,应用自回归移动平均季节乘积模型(autoregressive integrated moving average)进行预测。方法收集济南市2014—2020年两家食源性疾病主动监测哨点医院报告系统中的发病数据建立时间序列,采用SARIMA模型拟合发病情况,利用2021年的发病人数与预测值进行比较验证模型,并评价预测效果。结果建立SARIMA(2,0,1)(0,1,1)12模型较好的拟合了以往济南市食源性疾病的时间序列,AIC=687.22,使用Ljung-Box函数,获得了P=0.499,提示残差属于白噪声序列,2021年数据以检验模型外推效果较好,且实际值均落在预测值的95%的置信区间,模型预测效果比较理想。结论SARIMA(2,0,1)(0,1,1)12模型能够较好的拟合食源性疾病的时间变化,因此可用于食源性疾病月发病数的拟合和预测。Objectives To explore the trend characteristics of foodborne diseases in Jinan City and apply the seasonal autoregressive integrated moving average model(SARIMA)for prediction.Methods The incidence data of foodborne diseases from two active monitoring sentinel hospitals in Jinan City from 2014 to 2020 were collected to establish a time series.The SARIMA model was used to fit the incidence situation.The numbers of cases in 2021 were compared with the predicted values to validate the model and evaluate the predictive effect.Results The SARIMA(2,0,1)(0,1,1)12 model was established and fitted the time series of food borne diseases in Jinan well,with AIC=687.22.Using Ljung Box function,P=0.499 was obtained,indicating that the residual error belonged to the white noise series.The data in 2021 was used to test the model extrapolation effect,and the actual values fell within the 95%confidence interval of the predicted value.The model prediction effect was relatively ideal.Conclusion SARIMA(2,0,1)(0,1,1)12 model can better fit the temporal change of foodborne diseases,and therefore can be used to fit and predict the monthly incidence of foodborne diseases.

关 键 词：食源性疾病 ARIMA乘积季节模型 预测 

分 类 号：R181[医药卫生—流行病学]