证明了左椭球分布的极大似然估计性质.若随机矩阵Yn×p的分布是参数为μ,∑的左椭球分布,且Y^1,Y^2,…,Y^m,i·i·d,则μ,∑的极大似然估计为Y,S^2的充分必要条件是Y^i~Nn×p(μ,∑).
The left ellipsoidal distribution is derived from the method of the likelihood estimations of parrameters. Let Y be an n × p random matrix. If Y has left ellipsoidal distribution and Y^1 ,Y^2 ,…,Y^mi · i · d, the likelihood estimations of parrameters for μ,∑ are Y and S^2 ,if and only if Y^i-Nn×p (μ,∑).