| 无偏的岭回归迭代算法 |
| |
| 引用本文： | 葛宏立,方陆明.无偏的岭回归迭代算法[J].数学的实践与认识,1997(4). |
| |
| 作者姓名： | 葛宏立 方陆明 |
| |
| 作者单位： | 林业部华东设计院，浙江林学院计算中心 金华 321001，临安 311300 |
| |
| 摘 要： | 本文探讨线性模型的无偏的岭回归迭代算法,这种算法保持最小二乘法的性质,当存在较为严重的共线性时,它能给出较为精确的参数及其协差阵的估计值;当存在严格的共线性时,给出参数及其协差阵的无穷多解中的一个,这个解由初值决定。文章还给出了算法的收敛性及一些其它性质的证明。
|
| 关 键 词： | 线性回归模型 岭回归 共线性 |
An Unbiased Iteration Method Based on the Ridge Regression |
| |
| GE HONGLI.An Unbiased Iteration Method Based on the Ridge Regression[J].Mathematics in Practice and Theory,1997(4). |
| |
| Authors: | GE HONGLI |
| |
| Abstract: | Based on the ridge regression, an unbiased iteration method for estimating parameters and covariance matrixes of linear models is discussed. Its estimators keep all the properties of the LS estimators. It can give a relatively accurate estimate for a data set with serious collinearity or one solution of myriad, which depend on start values , when the data set is with strict collinearity. |
| |
| Keywords: | Linear model Ridge regression Collillearity Unbiased estimation |
| 本文献已被 CNKI 等数据库收录！ |
