摘要
对独立同分布样本情形的连续型单参数指数族的单边假设检验问题,在线性损失下导出了单调的Bayes检验函数,构造了相应的经验Bayes(EB)检验函数.在一定条件下,获得的经验Bayes检验函数的收敛速度可任意接近O(n^(-1)).最后给出了满足定理条件的两个例子.
In this paper, the empirical Bayes （EB） one-sided test problem for the continuous one-parameter exponential family is considered. The monotone Bayes test rule is derived and the EB test rule is constructed by using the independent and identically distributed （iid） samples. The convergence rate for the proposed EB test rule is obtained. It is shown that the convergence rates of the proposed EB test rules can arbitrarily close to O（n^-1） under suitable conditions. Finally, two examples concerning main result are given.
出处
《系统科学与数学》
CSCD
北大核心
2009年第8期1142-1152,共11页
Journal of Systems Science and Mathematical Sciences
基金
国家自然科学基金(10771204)
中国科学院知识创新工程重要方向项目(KJCX3-SYW-S02)资助
