上海交通大学学报(医学版)

• 论著(临床研究) • 上一篇    下一篇

2004—2012年中国丙型肝炎报告数据ARIMA模型及其趋势预测

吴田勇1,曾 庆1,于 萌2,刘世炜2,李 勤3,赵 寒3   

  1. 1.重庆医科大学 公共卫生学院卫生统计教研室, 重庆 400016; 2.中国疾病预防控制中心, 北京 102200; 3.重庆市疾病预防控制中心, 重庆 400042
  • 出版日期:2014-05-28 发布日期:2014-05-30
  • 通讯作者: 曾 庆, 电子信箱: zengqing1@gmail.com。
  • 作者简介:吴田勇(1987—), 男, 硕士生; 电子信箱: wutianyong1987@gmail.com。

ARIMA model of data of hepatitis C report of China from 2004 to 2012 and trend prediction

WU Tian-yong1, ZENG Qing1, YU Meng2, LIU Shi-wei2, LI Qin3, ZHAO Han3   

  1. 1.Department of Biostatistics, School of Public Health, Chongqing Medical University, Chongqing 400016, China; 2.Chinese Center for Disease Control and Prevention, Beijing 102200, China; 3.Chongqing Center for Disease Control and Prevention, Chongqing 400042, China
  • Online:2014-05-28 Published:2014-05-30

摘要:

目的 利用季节自动回归移动平均混合模型(ARIMA模型)对我国丙型病毒性肝炎(丙肝)报告数据进行分析、拟合和预测。方法 对2004年1月—2012年7月我国丙肝报告数据进行差分以达到平稳化,采用季节ARIMA模型对数据进行分析、拟合和预测。结果 2004年1月—2012年7月我国丙肝发病数呈逐年上升趋势,且呈现明显的以年为单位的周期性变化;对丙肝报告数据进行平稳化检验、差分、模型识别、模型诊断,获得季节序列ARIMA (1,1,1)×(1,1,1)12为最优模型,该模型残差检验为白噪声序列,且拟合数据在95%置信区间;对2012年7月―2014年12月全国丙肝发病数进行预测,显示全国丙肝发病数呈继续上升且具有明显的周期性波动趋势。结论 季节ARIMA模型能较好地拟合和预测我国丙肝发病数在时间上的变化趋势,可为疫情的防治提供借鉴。

关键词: 时间序列, 丙型肝炎, 季节ARIMA模型, 预测

Abstract:

Objective To analyze, fit, and forecast the data of hepatitis C report of China by the autoregressive integrated moving average model (ARIMA). Methods The data of hepatitis C report of China from January, 2004 to July, 2012 was differentiated to obtain smoothness and analysis, fitting, and forecasting were conducted by the seasonal ARIMA model. Results The incidence of Hepatitis C showed an increasing trend from January, 2004 to July, 2012 and has obvious periodic changes on a basis of one year. The smoothing test, differentiation, and model identification and diagnosis were conducted based on the data of Hepatitis C report and the obtained optimal model was seasonal series of ARIMA (1,1,1)×(1,1,1)12. The residual test of this model showed a white noise sequence, fitting the data in the 95% confidence interval. Based on the cases of hepatitis C from July, 2012 to December, 2014, it could be predicted that the incidence of hepatitis C increased continuously and showed an obvious trend of periodical fluctuation. Conclusion The seasonal series ARIMA model of time series can simulate and predict the tendency of incidence of hepatitis C in China and can provide reference for the prevention and control of the epidemic.

Key words: time series, hepatitis C, seasonal ARIMA model of seasonal, forecast