Communication-Efficient Feature Screening for Ultrahigh-Dimensional Data Under Quantile Regression

被引:0
|
作者
Diao, Tianbo [1 ,2 ]
Li, Bo [1 ]
机构
[1] Cent China Normal Univ, Sch Math & Stat, Wuhan, Peoples R China
[2] Nanyang Inst Technol, Sch Math & Sci, Nanyang, Peoples R China
来源
STAT | 2025年 / 14卷 / 02期
基金
中国国家自然科学基金;
关键词
distributed system; feature screening; sure screening property; ultrahigh-dimensional data; FEATURE-SELECTION;
D O I
10.1002/sta4.70055
中图分类号
O21 [概率论与数理统计]; C8 [统计学];
学科分类号
020208 ; 070103 ; 0714 ;
摘要
Feature screening is a powerful method for dealing with ultrahigh-dimensional data, but when the sample size is also increased, traditional methods encounter challenges due to constraints in data storage and computational complexity. In this article, we present a novel method for feature screening in quantile regression models implemented on the distributed system. To enhance convergence, we adopt an iterative projected gradient descent algorithm and establish the corresponding sure screening property. Extensive simulation studies confirm the effectiveness of the proposed method, and an application to a real dataset showcases its practical utility.
引用
收藏
页数:13
相关论文
共 50 条
  • [1] Quantile-Composited Feature Screening for Ultrahigh-Dimensional Data
    Chen, Shuaishuai
    Lu, Jun
    MATHEMATICS, 2023, 11 (10)
  • [2] Conditional quantile screening in ultrahigh-dimensional heterogeneous data
    Wu, Yuanshan
    Yin, Guosheng
    BIOMETRIKA, 2015, 102 (01) : 65 - 76
  • [3] SPARSE COMPOSITE QUANTILE REGRESSION WITH ULTRAHIGH-DIMENSIONAL HETEROGENEOUS DATA
    Qu, Lianqiang
    Hao, Meiling
    Sun, Liuquan
    STATISTICA SINICA, 2022, 32 (01) : 459 - 475
  • [4] A selective overview of feature screening for ultrahigh-dimensional data
    JingYuan Liu
    Wei Zhong
    RunZe Li
    Science China Mathematics, 2015, 58 : 1 - 22
  • [5] A selective overview of feature screening for ultrahigh-dimensional data
    Liu JingYuan
    Zhong Wei
    Li RunZe
    SCIENCE CHINA-MATHEMATICS, 2015, 58 (10) : 2033 - 2054
  • [6] A selective overview of feature screening for ultrahigh-dimensional data
    LIU JingYuan
    ZHONG Wei
    LI RunZe
    Science China(Mathematics), 2015, 58 (10) : 2033 - 2054
  • [7] Regularized quantile regression for ultrahigh-dimensional data with nonignorable missing responses
    Xianwen Ding
    Jiandong Chen
    Xueping Chen
    Metrika, 2020, 83 : 545 - 568
  • [8] Regularized quantile regression for ultrahigh-dimensional data with nonignorable missing responses
    Ding, Xianwen
    Chen, Jiandong
    Chen, Xueping
    METRIKA, 2020, 83 (05) : 545 - 568
  • [9] Efficient feature screening for ultrahigh-dimensional varying coefficient models
    Chen, Xin
    Ma, Xuejun
    Wang, Xueqin
    Zhang, Jingxiao
    STATISTICS AND ITS INTERFACE, 2017, 10 (03) : 407 - 412
  • [10] Communication-efficient estimation of high-dimensional quantile regression
    Wang, Lei
    Lian, Heng
    ANALYSIS AND APPLICATIONS, 2020, 18 (06) : 1057 - 1075