ARIMA乘积季节模型在全国布鲁菌病发病预测中的应用

Application of multiple seasonal ARIMA model in forecasting incidence of brucellosis in China

目的应用求和(差分)自回归移动平均(Autoregressive Integrated Moving Average Model,ARIMA)乘积季节模型分析季节性时间序列,建立全国布鲁菌病发病数的预测模型。方法利用中国疾病预防控制中心2011年1月—2016年12月按月报告的布鲁菌病发病数历史疫情数据,采用最大似然法估计模型参数,模型定阶后,建立布鲁菌病按月发病数ARIMA乘积季节预测模型。结果非季节和季节移动平均参数分别为0.357 35、0.666 64,均P<0.05,AIC=911.337 2,SBC=917.569 8,均P<0.05,据此建立ARIMA(0,1,1)(0,1,1)12乘积季节模型,模型表达式荦荦12xt=(1-0.357 35B)(1-0.666 6412)εt,并开展全国布鲁菌病发病数的预测。结论 ARIMA(0,1,1)(0,1,1)12乘积季节模型可用于预测布鲁菌病的发病情况。

[Objective]To analyze the seasonal time series by using Autoregressive Integrated Moving Average Model（ARIMA）,establish a predictive model for number of brucellosis cases in China.[Methods]On the basis of data about monthly number of brucellosis cases reported by China Center for Disease Control and Prevention from January 2011 to December 2016, the maximum likelihood method was used to estimate the model parameters, and the multiple seasonal ARIMA model was established for monthly number of brucellosis in China after parameters were identified.[Results]Norseasonal and seasonal moving average coefficients was 0.357 35 and 0.666 64 respectively. all P〈0.05, AIC =911.337 2, SBC =917.569 8. Autocorrelation test for residuals of model was white-noise series, all P〈0.05. ARIMA（0,1,1）（0,1,1）12 was established, and the model equation was荦荦12 xt=（1-0.357 35 B）（1-0.666 64 B12）εt. The number of brucellosis cases in China was forecasted.[Conclusion]The multiple seasonal model of ARIMA（0,1,1）（0,1,1）12 can be used to forecast number of brucellosis cases.