混合样本下的经验似然估计  

Empirical likelihood estimator for mixing dependent samples

作　　者：施明华[1] 赵建中[1] 周本达[1] 

机构地区：[1]皖西学院应用数学数学学院,安徽六安237012

出　　处：《纯粹数学与应用数学》2012年第4期462-468,506,共8页Pure and Applied Mathematics

基　　金：国家自然科学资金(61075049);安徽省高校优秀人才基金(2009SQRZ189);安徽省教育厅自然科学研究项目(KJ2009B113)

摘　　要：研究强平稳φ混合随机变量序列均值的经验似然估计问题,利用拉格朗日乘子以及一些重要概率不等式讨论均值有限且方差不等于零的强平稳φ混合序列,并给出其总体均值和M-泛函统计推断以及置信区间.In this paper,under the condition of φ-mixing random variables,the empirical likelihood method is studied.We establish statistical inference and confidence intervals are obtained for the population mean and Mfunctional of {Xn} which be the strongly stationary φ-mixing random variable sequence with mean is constant and a non negative variance by the lagrange multiplier and some important probability inequalities.

关 键 词：φ混合 经验似然 强平稳 

分 类 号：O211.7[理学—概率论与数理统计]