摘要
对有偏估计中的广义岭型主成分估计的优良性进行了较深入的研究.证明了广义岭型主成分估计优于最小二乘估计的充要条件,并在此基础上对几类常见的有偏估计在均方误差(阵)条件下优于最小二乘估计的充要条件进行了拓展.
In the paper , the choiceness of the generalized ridge principal component estimation in biased estimation was studied .In the process, the necessary and sufficient condition that the generalized ridge principal component estimation is superior to least square estimation was proved .In the light of this demonstration , the necessary and sufficient condition of a few common biased estimations was expanded , that is, they are superior to least square estimation under the condition of mean square error .
出处
《佳木斯大学学报（自然科学版）》
CAS
2014年第6期939-940,943,共3页
Journal of Jiamusi University：Natural Science Edition
基金
国家自然科学基金(11371030)
关键词
广义岭型主成分估计
最小二乘估计
均方误差
有偏估计
generalized ridge principal component estimation
least square estimation
mean square er-ror
biased estimation
