通过对一阶二分量随机系数混料回归模型变式,转变成混合效应混料回归模型,从而得到其信息矩阵.对一阶二分量随机系数混料回归模型,求出其单纯形一中心设计的D-最优设计和G-最优设计.在此基础上,证明一阶二分量随机系数混料回归模型不具有D-最优设计和G-最优设计等价性.
For the first order mixture regression model with two-component random coefficient, we use the vari-ant method to obtain the mixed-effect mixture regression model. We study the D-optimum designs and G-opti-mum designs for first order mixture regression model with two-component random coefficient. We show that the D-optimum designs and G-optimum designs for this model are not equivalent.