应用ARIMA模型对大庆市布鲁菌病发病率预测分析

Application of ARIMA model to forecast the incidence of brucellosis in Daqing

目的通过拟合ARIMA模型来预测大庆市布鲁菌病的发病趋势,为大庆市制定有效的布鲁菌病防治策略提供理论依据。方法以大庆市2009年1月-2016年5月布鲁菌病的发病情况为例,拟合ARIMA模型,并预测2016年12月-2017年5月的布鲁菌病发病率。结果最终拟合的模型为ARIMA(1.1.1)(0.1.0)12,残差的序列为白噪声序列。预测值和实际值的平均相对误差为13.39%。预测的2016年6-11月布鲁菌病发病率与实际值相符。预测2016年12月-2017年5月的发病率为1.494/10万、0.915/10万、1.930/10万、2.315/10万、1.442/10万和2.666/10万。结论 ARIMA模型能相对准确地预测近期大庆布鲁菌病的发病趋势。

Objective Fitting the ARIMA model to predict the incidence trend of brucellosis in Daqing, provide the theoretical basis for the effective control strategy of brucellosis in Daqing. Methods Taking the incidence of brucellosis from January 2009 to May 2016 in Daqing as an example,the ARIMA model was fitted and the incidence of brucellosis between December 2016 and May 2017 was predicted. Results The established finally model was ARIMA（ 1.1.1 ） （0.1.0） 12 and the residual se- quence was a white noise sequence. The relative error in average was 13.39% between the forecasting value and the real value. The predicted incidence from June 2016 to November 2016 was consistent with the actual one. From June 2016 to November2016, predictive brucellosis incidence were： 1. 494/100 thousand, 0. 915/100 thousand, 1. 930/100 thousand, 2. 315/100 thousand, 1. 442/100 thousand and 2.666/100 thousand. Conclusion ARIMA model could predict the trend of brucellosis in Daqing recently.