摘要
证明了在一类广义岭型降维估计中,广义岭型主成分估计的方差和最小。
It is proved that combining generalized ridge and principal component estimate has a piece of optimality, that is, its sum of variance is minimum in the class of generalized reduce dimension estimates.
出处
《青海大学学报（自然科学版）》
1999年第1期31-35,共5页
Journal of Qinghai University(Natural Science)
关键词
广义岭型
主成分估计
方差和
特征值
降维估计
Generalized ridge and component estimate, Sum of variance, Eigenvalue
