论文部分内容阅读
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。
The Autoregressive integrated moving average (ARIMA) model was used to predict the number of severe patients reported in our country (excluding Hong Kong, Maucao, Taiwan, China) in the monthly report of hand-foot-mouth disease. This model is suitable for prevention and control of HFMD and other infectious diseases The application provides the reference basis. According to the monthly report of HFMD in China from 2010 to 2015, the ARIMA model of HFRS reported in China was established based on the monthly reports of critically ill patients from January to September in 2016, 2010-2014 data model to be compared. The monthly numbers of severe patients reported by China in two different time series 2010-2014 and 2010-2015 were ARIMA (1,1,0) (2,1,0) 12, ARIMA (0,1, 1) (2,1,0) 12. The comparison of the prediction results of the above two different time series shows that more data are accumulated and the average relative error of the prediction is smaller, but the shorter the prediction time, the smaller the average relative error is found. In the same research content and time series, the forecasting models may be different. It is considered that the more ARIMA model data are accumulated, the shorter the forecasting time and the smaller the forecasting error need to be further verified.