摘要
在回归系数的主相关估计基础上提出一种新的降维估计——广义岭型主相关估计,并讨论它的优良性.
A new reduced-dimension estimation was proposed. It was called combining generalized ridge and principal correlation estimate in the base of the principal correlation estimate, and its optimality was discussed.
出处
《大学数学》
2009年第6期72-74,共3页
College Mathematics
关键词
回归模型
回归系数
广义岭型主相关估计
均方误差
regression model
regression coefficient
combining generalized ridge and principal correlation mean square error