自回归移动平均模型在恙虫病预测中的应用研究

Application of autoregressive moving average model in prediction of tsutsugamushi disease

目的探索基于季节性差分的自回归移动平均模型(ARIMA模型)在恙虫病预测应用的可行性。方法搜集中国疾病预防控制信息系统中的恙虫病发病资料,应用SPSS 17.0软件中的ARIMA模型,对北京市平谷区2010-2015年的恙虫病病例发病时间建立模型并拟合,根据模型对2016年的发病数做出预测。结果北京市平谷区恙虫病发病呈现逐年上升趋势,具有明显的季节性和周期性,每年的10月为发病高峰,经选取最优模型为ARIMA(1,2,2)(2,1,0)12,其平稳的R2=0.889,BIC=5.460,Ljung-Box Q检验,P=0.428,残差序列为白噪声序列。结论利用监测数据建立时间序列是预测传染病发展趋势的一个重要手段,此次建立的ARIMA模型对北京市平谷区恙虫病发病值及预测值拟合较好,可以作为恙虫病短期发病预测手段。

Objective To explore the feasibility of the application of autoregressive moving average model （ARIMA model）based on seasonal difference in prediction of tsutsugamushi disease. Methods The incidence of tsutsugamushi disease in Chinese disease pro- vention and control information system was collected and the incidence of tsutsugamushi disease in Pinggu District of Beijing was mod- eled and fitted by using the ARIMA model of SPSS 17. 0 software. According to the medel, incidence to make predictions. Results The incidence of tsutsugamushi disease in Pinggu District of Beijing showed a trend of increasing year by year, which was obviously seasonal and cyclical. The peak of the disease was in October, and the optimal model was ARIMA（1, 2, 2）（2, 1, O）12, the steady R- =0. 889, BIC =5. 460, Ljung- Box Q test, P =0. 428, the residual sequence is white noise sequence. Conclusion The establish- ment of time series by using monitoring data is an important means to predict the development trend of infectious diseases. The ARIMA model established in this paper is suitable for predicting the incidence and predictive value of tsutsugamushi disease in Pinggu District of Beijing, and can be used as a short -term predictive method for tsutsugamushi disease.