Auto-regressive模型在全国婴儿死亡率拟合中的应用  

Application of Auto-regressive Model in the National Infant Mortality Rate

作　　者：刘松[1] 李晓妹[1] 刘健[1] 刘晓冬[1] 李向云[1] 

机构地区：[1]潍坊医学院预防医学系,261053

出　　处：《中国卫生统计》2011年第4期366-368,共3页Chinese Journal of Health Statistics

摘　　要：目的分析我国1991～2007年的婴儿死亡率的变化规律,探讨Auto-regressive模型在非平稳时间序列数据拟合中的适用性和有效性。方法对我国婴儿死亡率数据序列的平稳性和纯随机性进行预处理,然后利用SAS程序拟合Auto-regressive模型,并根据决定系数R2评价其拟合效果。结果我国婴儿死亡率为非平稳时间序列,总体呈现随时间线性递减的长期趋势,同时又包含一定的随机信息,采用Auto-regressive模型拟合效果较好。结论 Auto-regressive模型可以用来拟合我国婴儿死亡率的数据,并可以推广应用到卫生领域中其他具有非平稳时间序列特征的数据,为相关卫生管理部门制定策略措施提供科学的理论依据。Objective To analyze the change of the national infant mortality rate between 1991 and 2007,and explore the applicability and effectiveness of Auto-regressive model in fitting non-stationary time series data.Methods We made pretreatment of time-series of national infant mortality rate to test the Stationarity and randomness,then used the SAS program fitting Auto-regressive model and evaluated the fitting results based on the coefficient of determination R2.Results Being non-stationary time series,there is a linear decreasing long-term trend over time,while it also contains random information.Auto-regressive model is suitable for fitting the national infant mortality rate.Conclusion Auto-regressive model can be used to fit the data of the national infant mortality rate,and can be applied to the field of health promotion in other non-stationary time series characteristics of data.It can provide a scientific theory for the relevant health authorities to develop strategies and measures.

关 键 词：Auto-regressive模型 婴儿死亡率 拟合 

分 类 号：R174[医药卫生—妇幼卫生保健]