ARIMA模型在上海市猩红热发病率预测中的应用

目的 探讨差分自回归移动平均(autoregressive integrated moving average,ARIMA)模型在上海市猩红热月发病率预测的应用。方法 利用ARIMA时间序列模型拟合2004年1月—2017年6月上海市猩红热的月发病率资料,并利用最优模型对2017年7—12月猩红热的月发病率进行预测。结果 最终拟合ARIMA(1,1,0)(0,1,1)₁₂模型,其标准化贝叶斯信息准则值(Bayesian information criterion,BIC)(-2.247)最小,残差经Ljung-Box Q(18)检验为白噪声序列,预测值与实际值基本吻合,相对误差在0.35%~16.74%的范围内。结论 ARIMA模型用于上海市猩红热月发病率的短期预测,可应用于定量风险评估等猩红热疫情的预警预测。

Application of ARIMA model to predicting the incidence rate of scarlet fever in Shanghai

Objective To explore the application of autoregressive integrated moving average (ARIMA) model to forecasting the monthly incidence rate of scarlet fever in Shanghai. Methods The ARIMA model was used to fit the monthly incidence rates of scarlet fever in Shanghai from January 2004 to June 2017, and the optimal model was used to predict the monthly incidence rates of scarlet fever in July-December 2017. Results ARIMA (1,1,0)(0,1,1)₁₂ was the optimal model, and the standardized Bayesian information criterion value was the smallest (-2.247). The residual was white noise sequence tested by Ljung box Q (18). The predicted value was basically consistent with the actual value, and the relative error was in the range of 0.35%-16.74%. Conclusions The ARIMA model can be used for short-term prediction of the incidence rate of scarlet fever in Shanghai, and can be applied to early warning prediction of scarlet fever epidemics, such as quantitative risk assessment.