半函数部分线性模型的经验似然推断  

Empirical Likelihood Inference for Semi-functional Partial Linear Model

作　　者：胡玉萍[1,2] 薛留根 冯三营[2] HU Yuping;XUE Liugen;FENG Sanying(College of Applied Sciences , Beijing University of Technology, Beijing 100024, China;School of Mathematics and Statistics , Zhengzhou University, Zhengzhou 450001 , China)

机构地区：[1]北京工业大学应用数理学院,北京100024 [2]郑州大学数学与统计学院,河南郑州450001

出　　处：《郑州大学学报（理学版）》2017年第4期5-10,共6页Journal of Zhengzhou University:Natural Science Edition

基　　金：国家自然科学基金项目(11501522;11571025;11331011);北京市自然科学基金项目(1142003;L140003);郑州大学青年启动基金项目(1512315004);郑州大学优秀青年基金项目(32210452)

摘　　要：考虑了半函数部分线性回归模型的估计问题.在函数型数据下发展了经验似然方法.构造了参数分量的经验似然比函数,得到提出的经验对数似然比渐近于χ~2分布,可用此结果构造兴趣参数的置信域.同时,也给出了非参数函数的估计,在一定的正则条件下给出了其收敛速度.The estimation for semi-functional partial linear model was considered. The empirical l ike l i-hood method was developed to make inference for parameter of interest. The empirical likelihood ratio for the parameter was constructed, and it was asymptotically standard chi-square distribution. Therefore, the corresponding confidence region of the parameter was constructed. At the same time, the estimator of the nonparametric function was given, and the theoretical property of the convergence rate was studied under certain regular conditions.

关 键 词：半函数部分线性模型 经验似然 函数型数据 置信域 

分 类 号：O212.7[理学—概率论与数理统计]