ARIMA预测模型在甘肃省其他感染性腹泻发病预测中的应用

Application of ARIMA prediction model in forecasting the incidence of other infectious diarrhea in Gansu Province

目的建立并评价甘肃省其他感染性腹泻发病的ARIMA预测模型。方法利用2010—2018年甘肃省其他感染性腹泻的发病数据建立ARIMA预测模型,同时利用2019年发病数据评价模型并对2020年甘肃省其他感染性腹泻发病进行预测。结果根据模型拟合效果,模型ARIMA(0,1,1)(1,1,0)12为最优模型。R^2=0.741,Ljung-Box检验值为25.944,BIC值为11.060。模型拟合甘肃省其他感染性腹泻的发病趋势与实际发病趋势一致,MAPE=17.297%,预测结果显示2020年甘肃省其他感染性腹泻发病时间分布与往年趋于一致。结论ARIMA(0,1,1)(1,1,0)12模型能较好地拟合甘肃省其他感染性腹泻的发病趋势,对该病的预防控制、风险评估等具有一定的公共卫生意义。

Objective To establish an autoregressive integrated moving average(ARIMA)prediction model in predicting the incidence of other infectious diarrhea in Gansu Province,and to evaluate its prediction effect.Methods Data regarding the incidence of other infectious diarrhea in Gansu Province from 2010 to 2018 were used to establish an ARIMA prediction model.At the same time,data about the incidence of other infectious diarrhea in 2019 were applied to evaluating the model,and the incidence of other infectious diarrhea in Gansu Province in 2020 was forecasted.Results The ARIMA(0,1,1)(1,1,0)12 model was supposed to be the best fitted model.The value of R^2 was 0.741,the value of Ljung-Box Q statistic was 25.944,and BIC value was 11.060.The model fitted the incidence trend of other infectious diarrhea in Gansu Province,which was consistent with the actual incidence trend,and the mean absolute percentage error(MAPE)was 17.297%.The predicted results revealed that the time distribution of incidence of other infectious diarrhea in Gansu Province in 2020 tended to be similar to that of previous years.Conclusions The data predicted by the ARIMA(0,1,1)(1,1,0)12 model can fit well with the incidence trend of other infectious diarrhea in Gansu Province,and the model has certain public health significance in prevention,control and risk assessment of other infectious diarrhea.