我国手足口病重症患者数自回归移动平均模型预测研究

Study on the National Monthly Reported Severe Cases of Hand-foot-mouth Disease Forecasted by Autoregressive Integrated Moving Average Model

采用自回归移动平均(Autoregressive integrated moving average,ARIMA)模型对我国(不含中国港澳台)手足口病月报告的重症患者数进行预测研究,为该模型在手足口病及其它传染病预防控制中的应用提供参考依据。根据2010-2015年全国手足口病月报告重症患者数时间序列,以2016年1-9月的月报告重症患者数作为验证数据,建立我国手足口病月报告重症患者数的ARIMA模型,并与2010-2014年数据建立的模型进行比较。2010-2014、2010-2015年两个不同时间序列建立的我国手足口病月报告重症患者数模型分别为ARIMA(1,1,0)(2,1,0)12、ARIMA(0,1,1)(2,1,0)12。以上两个不同时间序列预测结果比较发现,数据积累较多,预测的平均相对误差变小,但预测时间越短尚未发现平均相对误差较小。同一研究内容,时间序列年代不同,所建立的预测模型可能不同;认为ARIMA模型数据积累越多、预测时间越短、预测误差越小的情况还需得到进一步验证。

Autoregressive integrated moving average （ARIMA） model was used to predict the monthly re- ported severe cases of hand-foot-mouth disease（HFMD） in China to provide a reference for prevention and control of HFMD and the application of ARIMA in of ARIMA in HFMD and other infectious diseases. On the basis of time series supplied by the monthly reported severe cases of the national HFMD from 2010- 2015, ARIMA model was established with the actual cases of HFMD from January to September 2016 as the validated data and with the comparison of ARIMA model based on the data from 2010-2014 . The models based on the 2010-2014 and 2010-2015 data of monthly reported severe cases of HFMD in China areARIMA（1,1,0,）（2,1,0）12,ARIMA（0,1,1,）（2,1,0）12respectively. The comparison of two models shows that the average of the relative error decreases with the accumulated data and does not do the same with the shorter time of predication. Different time series may have different ARIMA models as for the same infectious diseases. It is needed to be further verified that the more data are accumulated, the more shorter time of predication is, the more smaller the average of the relative error is.