矩阵损失下线性预测关于不等式约束的容许性（英文）  

Admissibility of Linear Predictors under Matrix Loss Function with Respect to an Inequality Constraint

作　　者：何道江[1,2] 贺磊[2] 

机构地区：[1]北京理工大学数学系,北京100081 [2]安徽师范大学数学计算机科学学院,安徽芜湖241000

出　　处：《安徽师范大学学报（自然科学版）》2015年第4期322-326,共5页Journal of Anhui Normal University(Natural Science)

基　　金：The key project of distinguish Yong Talents of Universities in Anhui Province(2012QRL028ZD)

摘　　要：本文研究了有限总体模型中参数受到不等式r'β≥0约束时的容许性问题,在矩阵损失下得到了线性预测是可容许的充要条件,所给条件易于验证,便于应用.In this paper,we investigate the admissibility of linear predictors in a finite population model with respect to the inequality constraint r'β ≥ 0. Necessary and sufficient conditions for a linear predictor to be admissible in the class of linear predictors are obtained under matrix loss function. In addition,the conditions are demonstrated to be much easier to verify and to apply via an example.

关 键 词：有限总体 线性预测 矩阵损失 不等式约束 可容许性 

分 类 号：O151.21[理学—数学]