郑代坤, 谭毅, 李佳, 王军, 马帅, 沈忠周. 基于自回归求和移动平均模型预测我国手足口病月报告发病数[J]. 疾病监测, 2018, 33(1): 54-58. DOI: 10.3784/j.issn.1003-9961.2018.01.013
引用本文: 郑代坤, 谭毅, 李佳, 王军, 马帅, 沈忠周. 基于自回归求和移动平均模型预测我国手足口病月报告发病数[J]. 疾病监测, 2018, 33(1): 54-58. DOI: 10.3784/j.issn.1003-9961.2018.01.013
Zhen Daikun, Tan Yi, Li Jia, Wang Jun, Ma Shuai, Shen Zhongzhou. Prediction of monthly reported cases of hand foot and mouth disease by ARIMA model in China[J]. Disease Surveillance, 2018, 33(1): 54-58. DOI: 10.3784/j.issn.1003-9961.2018.01.013
Citation: Zhen Daikun, Tan Yi, Li Jia, Wang Jun, Ma Shuai, Shen Zhongzhou. Prediction of monthly reported cases of hand foot and mouth disease by ARIMA model in China[J]. Disease Surveillance, 2018, 33(1): 54-58. DOI: 10.3784/j.issn.1003-9961.2018.01.013

基于自回归求和移动平均模型预测我国手足口病月报告发病数

Prediction of monthly reported cases of hand foot and mouth disease by ARIMA model in China

  • 摘要: 目的 建立适合预测我国手足口病月报告发病人数的自回归求和移动平均(ARIMA)乘积季节模型,并评价其预测效果。方法 收集2010年3月至2017年7月我国手足口病月发病报告人数资料。通过R软件使用2010年3月至2017年1月的数据建立ARIMA乘积季节模型,并用2017年2-7月手足口病月发病报告人数评估该模型的预测效果,并对2017年8-12月的数据进行预测。结果 我国手足口病月发病报告数呈明显的周期性,且以24个月为一个周期重复,不具有长期趋势;建立了ARIMA(1,0,1)(0,1,1)24模型对我国手足口病月发病报告数进行预测;通过将预测数据与实际数据相比较,该模型预测绝对误差的平均值和相对误差的平均值分别为22 505.47和15.71%。结论 基于本研究的数据,ARIMA(1,0,1)(0,1,1)24模型可以拟合我国手足口病的月报告发病人数,可用于预测;同时也可为我国制定手足口病方面的防控措施以及评价防控效果提供科学的参考依据。

     

    Abstract: Objective To establish an autoregressive integrated moving average(ARIMA)model for the prediction of hand foot and mouth disease(HFMD)incidence in China and evaluate its forecosting ability. Methods The ARIMA model was established by using the incidence data of HFMD in China from March 2010 to January 2017 with software Excel 2007 and the incidence data of HFMD from February to July 2017 was used to evaluate the prediction abiliy of the model. Results The incidence of HFMD had obvious periodicity in China,i.e. 24 months. The ARIMA(1, 0, 1)(0, 1, 1)24 model was established to predict the incidence of HFMD in China. Compared with actual data from February to July 2017,the mean of absolute error and the relative error were 22 505.47 and 15.71%,respectively. Conclusion Based on the results of this study,the ARIMA(1, 0, 1)(0, 1, 1)24 model can be used for the fitting of HFMD incidence in China. It can not only predict the case number but also provide reference for the development of evaluation of prevention and control measures.

     

/

返回文章
返回