| 均匀设计实验资料最优回归子集的构建 |
| |
| 引用本文： | 胡良平,李子建,刘惠刚,张天明. 均匀设计实验资料最优回归子集的构建[J]. 中国医院统计, 2004, 11(4): 315-318 |
| |
| 作者姓名： | 胡良平 李子建 刘惠刚 张天明 |
| |
| 作者单位： | 1. 100850 北京军事医学科学院生物医学统计咨询中心
2. 首都医科大学基础医学院 |
| |
| 基金项目： | 军事医学科学院院创新项目 |
| |
| 摘 要： | 目的本文针对均匀设计实验资料的回归分析结果不唯一的事实,提供构建均匀设计实验资料最优回归子集的策略.方法拟合包含不同数目自变量的回归方程,运用SAS语言编写的程序对方程进行假设检验以获得最大F值为判断标准寻找最优回归方程,并用一个真实的实验资料演示这一计算过程和结果.结果能够对给定的均匀设计实验资料以较快的速度和较高的精确度获得"较优(甚至最优)"的回归方程.结论本文所提供的方法是构建均匀设计实验资料最优回归子集的一个有效方法.
|
| 关 键 词： | 均匀设计 最优回归子集 多元线性回归方程 |
| 文章编号： | 1006-5253(2004)04-0315-04 |
| 修稿时间： | 2004-08-17 |
Creating Best Regression Subsets for the Experimental Data of the Uniform Design |
| |
| Hu Liangping,Li Zijian,LiuHuigang,et al.. Creating Best Regression Subsets for the Experimental Data of the Uniform Design[J]. Chinese Journal of Hospital Statistics, 2004, 11(4): 315-318 |
| |
| Authors: | Hu Liangping Li Zijian LiuHuigang et al. |
| |
| Affiliation: | Hu Liangping,Li Zijian,LiuHuigang,et al. Consulting Center of Biomedical Statistics,Academy of Military Medical Sciences,Beijing 100850,China, Academy of Continued Education,Capital University of Medical Sciences,Beijng 100036,China |
| |
| Abstract: | Objective Based on the fact that the regression analysis results of the data collected from the Uniform Design are not unique,the author supplied a strategy of creating best regression subsets for the data. Methods A series of regression equations with different number independent variables are fitted, so that best regression subsets can be obtained according to the criteria of maximum F value by using SAS language. Furthermore, the calculating procedure and results are demonstrated by analyzing real experimental data. Results For the given data collected from the Uniform Design, a well -fitted (or best) regression equation can be obtained rapidly and accurately. Conclusion The method provided in this paper is an effective approch to creating best regression subsets for the data of the Uniform Design. |
| |
| Keywords: | Uniform design Best Regression subset Multivariate linear regression equation |
| 本文献已被 CNKI 维普 万方数据 等数据库收录！ |
|
