文章摘要
ARIMA季节乘积模型在新疆地区细菌性痢疾发病率预测中的应用
Application of Multiple Seasonal ARIMA Model on Forecasting Incidence of Bacillary Dysentery in Xinjiang
投稿时间:2019-07-24  修订日期:2019-07-24
DOI:
中文关键词: 时间序列分析  求和自回归移动平均模型  细菌性痢疾  预测
英文关键词: time series analysis  multiple seasonal ARIMA model  bacillary dysentery  forecasting
基金项目:国家自然科学基金:基于周期数据的广义保形拟插值的构造理论及其应用。项目编号:11501006
作者单位E-mail
沈彭 安徽大学 1083617848@qq.com 
魏峰 安徽大学 2326372769@qq.com 
摘要点击次数: 10
全文下载次数: 
中文摘要:
      探讨ARIMA季节乘积模型在新疆地区细菌性痢疾疫情预测预警中应用的可行性,为细菌性痢疾的防控工作提供科学依据。方法 对新疆2004-2016年细菌性痢疾月发病率数据进行分析,建立ARIMA季节乘积模型(p,d,q)(P,D,Q)s;并利用2016年5~12月的数据进行模型预测效果的评价。结果 新疆2004年1月至2016年4月细菌性痢疾报告发病率有明显的季节性变动趋势;通过数据差分、模型拟合、评价等过程,获得的相对最佳模型为ARIMA(0,0,1) (0,1,0)12;该模型残差序列为白噪声序列。结论 ARIMA季节乘积模型能够有效的预测新疆地区细菌性痢疾发病率的短期变化趋势。
英文摘要:
      Objective: This study aimed to discuss the feasibility on forecasting incidence of bacillary dysentery with the multiple seasonal ARIMA model in Xinjiang; thus to provide a scientific basis for future prevention and control of bacillary dysentery. Methods: The ARIMA (p, d, q) (P, D, Q) <sub>s</sub> model was used to fit the data about monthly incidence of bacillary dysentery in Xinjiang from January 2004 to April 2016; and its forecasting precision was evaluated by using the data from Mar 2016 to Decembe 2016. Results: The monthly incidence of bacillary dysentery in Xinjiang from January 2004 to April 2016 showed a seasonal change in trend; after data differencing, model fitting and diagnosing, the ARIMA (0, 0, 1) (0, 1, 0) <sub>12</sub> model was found as the best relatively optimal model for fitting the data; and the residuals of this model were tested as white noise sequences. Conclusions: The results showed that the multiple seasonal ARIMA model could predict the incidence of bacillary dysentery in Xinjiang very effectively in short-term.
View Fulltext   查看/发表评论  下载PDF阅读器
关闭

手机扫一扫看 分享按钮