摘要
本文目的是通过两个实例介绍适应性回归样条算法对具有混杂结构的数据集进行回归建模的实践。当一个数据集包含多个不同的回归模型时,只要给定分类变量的具体取值,SAS中的"ADAPTIVEREG过程"可以比较准确地发掘出其内在规律,并以图形方式呈现模型对资料的拟合效果,还能呈现由"基函数"及其组合所构造出的"回归模型"。图形呈现的结果确实可以给分析者提供一些有价值的"分析线索"或"积极暗示",缩小"探索性研究的空间"。
This paper was to introduce the application of the building regression model adopted the adaptive regression spline algorithm on the data sets with mixed structure through two examples. When a data set contained several distinct regression models, the ADAPTIVEREG procedure in SAS software could discover the internal rules as long as the specific values of the classification variables were given, and could present the fitted results of the model with graphs and the regression model constructed with the basis functions and their combinations. The information presented in the graphs could provide some valuable analytical clues or positive hints for the analysts, which greatly reduced the space of the exploratory research.
作者
罗艳虹
胡良平
Luo Yanhong;Hu Liangping(Department of Health Statistics,School of Public Health,Shanxi Medical University,Taiyuan 030001,China;Specialty Committee of Clinical Scientific Research Statistics of World Federation of Chinese Medicine Societies,Beijing100029,China;Graduate School,Academy of Military Sciences PLA China,Beijing 100850,China)
出处
《四川精神卫生》
2019年第2期105-109,共5页
Sichuan Mental Health
基金
国家高技术研究发展计划课题资助(2015AA020102)
关键词
混合结构
适应性
样条
回归分析
基函数
失拟
拟合优度
Mixed structure
Adaptability
Spline
Regression analysis
Basis function
Lack of fit
Goodness of fit
