| 非齐次等式约束线性回归模型回归系数的综合条件岭估计 |
| |
| 引用本文： | 农秀丽,彭展声.非齐次等式约束线性回归模型回归系数的综合条件岭估计[J].科技通报,2012,28(2):4-6,9. |
| |
| 作者姓名： | 农秀丽 彭展声 |
| |
| 作者单位： | 广西民族师范学院数学与计算机科学系,广西崇左,532200 |
| |
| 基金项目： | 广西自然科学基金项目(2010GXNSFA013116);广西高校优秀人才资助项目(桂教人〔2010〕65号);广西民族师范学院基金项目(ybxm200908) |
| |
| 摘 要： | 提出非齐次等式约束线性回归模型回归系数的一个新的有偏估计,即综合条件岭估计,讨论了综合条件岭估计的性质,在一定的条件下,综合条件岭估计的样本总方差、均方误差、均方误差矩阵均分别小于约束最小二乘估计的相应误差.在综合条件岭估计下,条件岭估计和条件根方估计为其特例,从而统一了条件岭估计和条件根方估计的理论.
|
| 关 键 词： | 回归模型 条件岭估计 条件根方估计 均方误差 均方误差矩阵 |
The Synthesized Conditional Ridge Estimation of Regression Coefficient in the Nonhomogeneous Equation Restricted Linear Regression Model |
| |
| NONG Xiuli
,
PENG Zhansheng.The Synthesized Conditional Ridge Estimation of Regression Coefficient in the Nonhomogeneous Equation Restricted Linear Regression Model[J].Bulletin of Science and Technology,2012,28(2):4-6,9. |
| |
| Authors: | NONG Xiuli PENG Zhansheng |
| |
| Institution: | (Department of Mathematics and Computer science GuangXi Normal University for Nationalities, Chongzuo 532200,China) |
| |
| Abstract: | To improve on the restricted least squares estimation(RLSE),A new biased estimation,the synthesized conditional ridge estimation is considered.For the new biased estimation,superiority over the RLSE is proved.Show that it is superior to RLSE in terms of covariance and mean squares error and mean squares error matrix.This estimation includes the conditional ridge estimation and root root estimation.Under the synthesized conditional estimation,the conditional ridge estimation and root root estimation are unified. |
| |
| Keywords: | regression model conditional ridge estimation root root estimation ean squares error mean squares error atrix |
| 本文献已被 CNKI 万方数据 等数据库收录！ |
