论文部分内容阅读
目的建立湖南省手足口病发病趋势的SARIMA模型,为手足口病的预防和控制提供参考依据。方法收集中国疾病预防控制信息系统2008年5月—2013年12月湖南省手足口病月发病率数据建模,以2014年1—7月的月发病率数据进行验证,并对2014年8月—2015年7月发病情况进行预测;应用SPSS 18.0中的“Define Dates”模块和“Fore Casting”模块进行分析,建立季节性差分自回归移动平均模型(SARIM A)。结果湖南省手足口病月发病率发病趋势预测模型为SARIMA(1,0,0)(1,1,0)12,模型自回归参数AR1=0.765,(t=8.789,P<0.001),残差为白噪声(Ljung-Box Q=15.420,P=0.494),预测值与实际值的相对误差范围为6.90%~46.31%,平均相对误差为20.37%;预测2014年发病率2次高峰分别在5月份和11月份,均高于2013年同月份的发病率;2015年上半年高峰期也在5月份,低于2014年同月份的发病率。结论 SARIM A(1,0,0)(1,1,0)12拟合效果较好,可用于湖南省手足口病月发病率的短期预测。
Objective To establish SARIMA model of HFMD in Hunan Province and provide a reference for the prevention and control of HFMD. Methods The data of monthly incidence of HFMD in Hunan Province from May 2008 to December 2013 were collected from China Disease Prevention and Control Information System and validated by monthly incidence data from January to July 2014. The data of August 2014 - The incidence of July 2015 was predicted. The “Define Dates” module and the “Fore Casting” module of SPSS 18.0 were used for analysis to establish the seasonal differential autoregressive moving average model (SARIM A). Results The predictive model for monthly incidence of HFMD in Hunan Province was SARIMA (1,0,0) (1,1,0) 12. AR1 was 0.765 (t = 8.789, P <0.001) The relative error between predicted and actual values ranged from 6.90% to 46.31% and the average relative error was 20.37%. The second-order peak incidence in 2014 was May and November were higher than the incidence in the same month in 2013; the first half of 2015 was also at the peak in May, lower than the incidence in the same month in 2014. Conclusion SARIM A (1, 0, 0) (1, 1, 0) 12 has a good fitting effect and can be used for the short-term prediction of monthly incidence of HFMD in Hunan Province.