THE CONSISTENCY FOR THE WEIGHTED ESTIMATOR OF NON-PARAMETRIC REGRESSION MODEL BASED ON WIDELY ORTHANT-DEPENDENT ERRORS

被引：11
|
作者：
Xia, Hao [1 ]
Wu, Yi [1 ]
Tao, Xinran [1 ]
Wang, Xuejun [1 ]
机构：
[1] Anhui Univ, Sch Math Sci, Hefei 230601, Anhui, Peoples R China
来源：
PROBABILITY IN THE ENGINEERING AND INFORMATIONAL SCIENCES | 2018年 / 32卷 / 03期
基金：
中国国家自然科学基金;
关键词：
complete consistency; convergence rate; non-parametric regression model; weighted estimator; widely orthant-dependent errors; FIXED-DESIGN REGRESSION; PRECISE LARGE DEVIATIONS; LINEAR-TIME SERIES; RANDOM-VARIABLES; MULTIPLE-REGRESSION; SEQUENCES;
D O I：
10.1017/S0269964817000262
中图分类号：
T [工业技术];
学科分类号：
08 ;
摘要：
In this paper, the complete consistency for the weighted estimator of non-parametric regression model based on widely orthant-dependent errors is established, where the restriction imposed on the dominating coefficient g(n) is very general. Moreover, under some stronger moment condition, we further obtain the convergence rate of the complete consistency, where the assumption on the dominating coefficient g(n) is also very general.
引用
收藏
页码：469 / 481
页数：13
相关论文
共 50 条
12345