共 50 条
Optimal designs for mixture models with amount constraints
被引:14
|作者:
Zhang, Chongqi
[1
]
Wong, Weng Kee
[2
]
机构:
[1] Guangzhou Univ, Dept Probabil & Stat, Guangzhou 510006, Guangdong, Peoples R China
[2] Univ Calif Los Angeles, Dept Biostat, Fielding Sch Publ Hlth, Los Angeles, CA 90095 USA
关键词:
A-optimal design;
Budget constraint;
D-optimal design;
Mixture-amount model;
Model intercept;
EQUIVALENCE;
D O I:
10.1016/j.spl.2012.08.029
中图分类号:
O21 [概率论与数理统计];
C8 [统计学];
学科分类号:
020208 ;
070103 ;
0714 ;
摘要:
Optimal designs for mixture experiments defined on the regular simplex are widely available but a lot fewer optimal designs are available for mixture experiments defined on the q-dimensional design space: S-*(q) = {(Z(1), . . . , Z(q))' is an element of R-q vertical bar Z(1) + . . . + Z(q) <= 1, Z(i) >= 0, i = 1, . . . , q}. This paper proposes a flexible class of models for mixture experiments defined on S-*(q) and gives a simple method for finding and A-optimal designs for the models using optimal designs available for the mixture experiments defined on the regular simplex. (C) 2012 Elsevier B.V. All rights reserved.
引用
收藏
页码:196 / 202
页数:7
相关论文