指数型分布族中矩估计的序贯置信区间  

Sequential Confidence Interval for Moment Estimate in the Exponential Distribution Family

作　　者：钱能生[1] 陈思宝[1] 

机构地区：[1]五邑大学数学物理系,广东江门529020

出　　处：《五邑大学学报（自然科学版）》2005年第2期1-6,共6页Journal of Wuyi University(Natural Science Edition)

摘　　要：在矩估计的基础上,对于给定精度(2d)及置信系数(α),建立了对参数(θ)函数g(θ)的一个序贯置信区间估计的步骤和方法.并讨论了在一定条件下,当d→0时,估计的渐近相合性、渐近有效性及有界的最优费用差(EN(d)?n(d))等渐近性质.The aim of this paper is to build a program of sequential confidence interval for the function of parameter g (θ ) under the conditions of prescribed precision (2d) and coverage probability (α ) based on moment estimation. Some of its asymptotic properties, such as asymptotic consistency, asymptotic efficiency and cost of ignorance under certain conditions, are also discussed.

关 键 词：序贯置信区间 渐近相合性 渐近有效性 最优费用差 矩估计 

分 类 号：O212.3[理学—概率论与数理统计]