基于指数平滑状态空间模型预测河南省猩红热发病趋势

Prediction of incidence trend of scarlet fever in Henan based on exponential smoothing state space model

目的探索指数平滑状态空间模型(ESSS)在河南省猩红热发病率预测中的应用,为未来猩红热防控工作提供科学依据。方法收集2004年1月至2018年12月河南省猩红热发病时间序列,使用2004年1月至2017年12月的数据建立ESSS模型和季节性求和自回归滑动平均混合模型(SARIMA),并将它们的拟合及预测性能进行比较。结果SARIMA(1,0,0)(0,1,1)12和ESSS(M,AD,M)模型被识别为最佳的拟合模型,并使用它们分别对2018年的月数据进行预测。比较发现,ESSS(M,AD,M)模型的拟合和预测值的均方根误差、平均绝对误差、平均误差率、平均绝对百分比误差和均方根百分比误差均小于SARIMA(1,0,0)(0,1,1)12模型,特别在预测方面,上述指标的值更小。结论ESSS模型在预测猩红热发病序列方面优于SARIMA模型,其可以作为一种有效的工具对猩红热未来的流行模式进行预测。

Objective To explore the suitability of exponential smoothing state space(ESSS)model for application in forecasting the incidence of scarlet fever in Henan province,in order to provide scientific evidence for the prevention and control of scarlet fever in the future.Methods Data on the incidence time series of scarlet fever from January 2004 to December 2018 was collected.The data from January 2004 to December 2017 was treated as the training set to develop the ESSS and seasonal autoregressive integrated moving average(SARIMA)models,and the remaining data were used for testing set to compare their predictive abilities between SARIMA model and ESSS model.Results SARIMA(1,0,0)(0,1,1)12 and ESSS(M,AD,M)were identified as the best-fitting SARIMA and ESSS models for modeling the scarlet fever incidence data in Henan province,and they were used to forecast the monthly incidence of scarlet fever in 2018.By comparing the evaluation criteria(including root mean square error,mean absolute error,mean error rate,mean absolute percentage error,and root mean square percentage error)between these two models,it was found that the ESSS model presented lower values for the above-mentioned indices,especially when it was used for prediction.Conclusion The ESSS model outperforms the most frequently used SARIMA model in estimating the incidence of scarlet fever both in fitting and forecasting aspects.It can be used as an effective tool to predict the future epidemic patterns of scarlet fever.