| 基于泛岭估计对岭估计过度压缩的改进方法 |
| |
| 引用本文： | 刘文卿.基于泛岭估计对岭估计过度压缩的改进方法[J].数理统计与管理,2011,30(4):614-619. |
| |
| 作者姓名： | 刘文卿 |
| |
| 作者单位： | 中国人＆大学统计学院,北京,100872 |
| |
| 摘 要： | 岭估计是解决多元线性回归多重共线性问题的有效方法,是有偏的压缩估计。与普通最小二乘估计相比,岭估计可以降低参数估计的均方误差,但是却增大残差平方和,拟合效果变差。本文提出一种基于泛岭估计对岭估计过度压缩的改进方法,可以改进岭估计的拟合效果,减小岭估计残差平方和的增加幅度。
|
| 关 键 词： | 泛岭估计 过度压缩 多重共线性 残差平方和 |
An Improved Method based Universal Ridge Estimate for the Excess Shrinkage of Ridge Estimate |
| |
| LIU Wen-qing.An Improved Method based Universal Ridge Estimate for the Excess Shrinkage of Ridge Estimate[J].Application of Statistics and Management,2011,30(4):614-619. |
| |
| Authors: | LIU Wen-qing |
| |
| Institution: | LIU Wen-qing (School of Statistics,Renmin Univercity of China,Beijing 100872,China) |
| |
| Abstract: | Ridge estimate is an effective method to solve the problem of multicoUinearity in multiple linear regression,and is a biased shrinkage estimate.Against ordinary least squares estimate,ridge estimate decreases mean square errors,but increases residual sum of squares.This paper proposes an improved method based universal ridge estimate for the excess shrinkage of ridge estimate.Which method can make better the effect of fit,reduces the residual sum of squares contrast to ridge estimate. |
| |
| Keywords: | universal ridge estimate excess shrinkage multicollinearity residual sum of squares |
| 本文献已被 CNKI 万方数据 等数据库收录！ |
|
