SARIMA模型和季节趋势模型在手足口病发病率预测中的应用

Application of SARIMA model and seasonal trend model to predict the incidence of hand,foot and mouth disease

目的比较自回归滑动平均求和季节乘积模型(SARIMA)和季节趋势模型对手足口病(HFMD)发病率的预测效能。方法利用荆州市2010—2015年的手足口病逐月发病率作为拟合数据,以2016年的逐月发病率作为预测数据,分别建立SARIMA模型和季节趋势模型后,根据2个模型的平均绝对百分比误差(MAPE)、平均误差率(MER)、均方误差(MSE)和平均绝对误差(MAE)评价模型的效果。结果 SARIMA(1,0,0)(1,1,0)12(不含常数项)拟合和预测的MAPE、MER、MSE、MAE为18.86%,25.47%,24.23,2.17和15.61%、16.71%、9.41、2.02;季节趋势模型拟合和预测的MAPE、MER、MSE和MAE为33.91%、39.48%、44.38、3.31和21.06%、20.32%、12.63、2.54。结论 SARIMA模型拟合效果较好,预测精度更高,可应用于手足口病疫情的预测和预警。

Objective To compare the efficiency of multiple seasonal autoregressive moving average(SARIMA)model and seasonal trend model to predict the incidence of hand,foot and mouth disease(HFMD)in Jingzhou city.Methods HFMD monthly incidence data from 2010 to 2015 in Jingzhou city were used to fit models,the data from Dec 2016 were used for prediction.SARIMA model and seasonal trend model were established,whose effects were evaluated by comparison of mean absolute percentage error(MAPE),Mean Error Rate(MER),mean square error(MSE)and mean absolute error(MAE).Results The MAPE,MER,MSE and MAE fitted and predicted by SARIMA(1,0,0)(1,1,0)12(no constant)model were18.86%,25.47%,24.23,2.17 and 15.61%,16.71%,9.41,2.02;respectively.The MAPE,MER,MSE and MAE fitted and predicted by seasonal trend model were 33.91%,39.48%,44.38,3.31 and 21.06%,20.32%,12.63,2.54;respectively.Conclusion SARIMA model could be fitted more effectively with high predictive accuracy,which can be used for prediction and warning of HFMD epidemics.