| 指示变量随机缺失下变系数模型的分位数回归 |
| |
| 引用本文： | 宁黎明,何晓霞,王志明. 指示变量随机缺失下变系数模型的分位数回归[J]. 武汉科技大学学报, 2019, 0(3): 227-234 |
| |
| 作者姓名： | 宁黎明 何晓霞 王志明 |
| |
| 作者单位： | 武汉科技大学理学院 |
| |
| 基金项目： | 国家自然科学基金资助项目(11201356); 湖北省教育厅人文社会科学基金资助项目(17Q044);武汉科技大学冶金工业过程系统科学湖北省重点实验室基金资助项目(201715). |
| |
| 摘 要： | 本文研究了删失数据以及删失指示量随机缺失情况下部分线性变系数模型的参数估计问题。对非参数部分采用B样条近似,对缺失的指示变量运用极大似然估计,结合分位数回归,得到了参数估计的渐近正态性质和非参数部分的收敛速度。蒙特卡洛模拟和实例分析表明,所提出的方法可以有效处理此类存在缺失的数据,获得有意义的结果。
|
| 关 键 词： | 部分线性变系数模型 分位数回归 删失数据 删失指示量 随机缺失 B样条 |
| 收稿时间： | 2019-01-05 |
Quantile regression of varying-coefficient models with indicators missing at random |
| |
| Ning Liming,He Xiaoxia and Wang Zhiming. Quantile regression of varying-coefficient models with indicators missing at random[J]. Journal of Wuhan University of Science and Technology, 2019, 0(3): 227-234 |
| |
| Authors: | Ning Liming He Xiaoxia Wang Zhiming |
| |
| Affiliation: | College of Science, Wuhan University of Science and Technology, Wuhan 430065, China,College of Science, Wuhan University of Science and Technology, Wuhan 430065, China and College of Science, Wuhan University of Science and Technology, Wuhan 430065, China |
| |
| Abstract: | This paper studies the parameter estimation of partially linear varying-coefficient model with censored data and censoring indicators missing at random. B-spline is used to approximate the non-parametric components and the maximum likelihood estimation is applied to the missing indicator variables. Based on quantile regression method, asymptotic normality of the parameter estimation and convergence speed of the non-parametric part are obtained. Monte Carlo simulations and real data analysis show that the proposed method can effectively process these censored data and acquire meaningful results. |
| |
| Keywords: | partially linear varying-coefficient model quantile regression censored data censoring indicator missing at random B-spline |
| 本文献已被 CNKI 等数据库收录！ |
| 点击此处可从《武汉科技大学学报》浏览原始摘要信息 |
|
点击此处可从《武汉科技大学学报》下载免费的PDF全文 |
