具有限制条件的部分线性模型的经验似然推断（英文）  

Empirical Likelihood Statistical Inference for Partially Linear Model with Restricted Condition

作　　者：刘常胜[1] 李永献[1] LIU Chang-sheng LI Yong-xian(Department of Mathematics and Physics, Henan University of Urban Construction, Pingdingshan 467000, Chin)

机构地区：[1]河南城建学院数理系,中国平顶山467000

出　　处：《湖南师范大学自然科学学报》2017年第4期75-82,共8页Journal of Natural Science of Hunan Normal University

基　　金：河南省科技计划项目资助(112300410191)

摘　　要：本文将经验似然方法应用到具有限制假设条件的部分线性模型中.为了检验假设条件,构造基于零假设和对立假设条件下的极大经验对数似然比估计值的差值统计量.而且在零假设下证明该统计量的极限分布为标准的χ~2分布.数值模拟表明所提出的检验统计量的优势.In this paper, we apply the empirical likelihood method to partially linear model with parameter linear restricted hypothesis.For the sake of testing hypothesis, an empirical log-likelihood ratio test statistic based on the difference of the null and alternative hypotheses is constructed.Furthermore, the limiting distribution of the test statistics is proved to be a standard Chi-squared distribution.Numerical simulation confirms the advantage of the proposed method.

关 键 词：经验似然 限制条件 部分线性模型 假设检验 Χ^2分布 

分 类 号：O212.1[理学—概率论与数理统计]