摘要
目的探讨自回归求和移动平均(ARIMA)模型在其他感染性腹泻发病预测中的应用,为其他感染性腹泻防控提供参考依据。方法应用SPSS20.0对2010年1月至2017年12月我国其他感染性腹泻月发病数构建ARIMA模型,以2018年各月月发病数评价模型的预测效果,并采用构建的最优模型进一步预测2019年我国其他感染性腹泻分月发病情况。结果构建的我国其他感染性腹泻发病预测模型为ARIMA(2,1,1)(0,1,1)12,每个参数均有统计学意义(P<0.05)。贝叶斯信息准则(BIC)=19.434,模型Box-Ljung(LB)检验差异无统计学意义(Q=9.006,P=0.831),故残差序列是白噪声序列,预测的平均相对误差为15.61%。模型短期预测效果较好,预计2019年我国其他感染性腹泻月平均发病数为112 985例,发病水平较2018年有总体上升趋势。结论ARIMA(2,1,1)(0,1,1)12模型对我国其他感染性腹泻发病拟合的预测效果较好,可用于其发病数的短期预测。预测结果显示2019年其他感染性腹泻发病有上升趋势,需加强其防控措施。
Objective To discuss the application of autoregressive integrated moving average(ARIMA)model in the prediction of the incidence of other infectious diarrhea,and provide references for the prevention and control of other infectious diarrhea.Methods The ARIMA model was established by using SPSS20.0 for the monthly incidence of other infectious diarrhea in China from January 2010 to December 2017.The predictive effect of the model was evaluated by the number of monthly epidemics from January to December 2018.The best model would be applied to predict the monthly incidence of other infectious diarrhea in China in 2019.Results The predictive models of other infectious diarrhea in China were ARIMA(2,1,1)(0,1,1)12.Each parameter was statistically significant(P<0.05).Bayesian information criteria(BIC)was 19.434,and the difference between the model Box-Ljung(LB)test was not statistically significant(Q=9.006,P=0.831).The residual sequence was white noise sequence,and the mean relative error was 15.61%.The short-term prediction effect of the model was better.It was estimated that the average monthly incidence of other infectious diarrhea in China in 2019 would be 112 985 cases,and the incidence level would be higher than that in 2018.Conclusion The ARIMA(2,1,1)(0,1,1)12 model has a good predictive effect for other infectious diarrhea in China,and can be used for short-term prediction of this disease.The predicted results showed that the incidence of other infectious diarrhea was on the rise in 2019,and the prevention and control measures should be strengthened.
作者
魏新刚
曹桂芝
李涵博
耿雪
郑贵森
WEI Xingang;CAO Guizhi;LI Hanbo;GENG Xue;ZHENG Guisen(School of Public Health,Gansu University of Chinese Medicine,Lanzhou,Gansu 730000,China;Shanghai Jiao Tong University School of Medicine,Shanghai 200025,China)
出处
《现代医药卫生》
2019年第21期3281-3284,共4页
Journal of Modern Medicine & Health