Data-driven slicing for dimension reduction in regressions:A likelihood-ratio approach

扫码查看

原文链接

万方数据
维普

外文摘要：To efficiently estimate the central subspace in sufficient dimension reduction,response discretization via slicing its range is one of the most used methodologies when inverse regression-based methods are applied.However,existing slicing schemes are almost all ad hoc and not widely accepted.Thus,how to define data-driven schemes with certain optimal properties is a longstanding problem in this field.The research described here is then twofold.First,we introduce a likelihood-ratio-based framework for dimension reduction,subsuming the popularly used methods including the sliced inverse regression,the sliced average variance estimation and the likelihood acquired direction.Second,we propose a regularized log likelihood-ratio criterion to obtain a data-driven slicing scheme and derive the asymptotic properties of the estimators.A simulation study is carried out to examine the performance of the proposed method and that of existing methods.A data set concerning concrete compressive strength is also analyzed for illustration and comparison.

外文关键词：

full-likelihood approachadaptive slicingregularizationsecond-order method

作者：

Peirong Xu、Tao Wang、Lixing Zhu

展开 >

作者单位：

Department of Statistics,Shanghai Jiao Tong University,Shanghai 200240,China

Center for Statistics and Data Science,Beijing Normal University at Zhuhai,Zhuhai 519087,China

基金：

国家自然科学基金国家自然科学基金Shanghai Rising-Star ProgramMultidisciplinary Cross Research Foundation of Shanghai Jiao Tong UniversityMultidisciplinary Cross Research Foundation of Shanghai Jiao Tong UniversityMultidisciplinary Cross Research Foundation of Shanghai Jiao Tong UniversityNeil Shen's SJTU Medical Research Fund of Shanghai Jiao Tong University

项目编号：

119710171197101820QA1407500YG2019QNA26YG2019QNA37YG2021QN06

出版年：

2024

DOI：

10.1007/s11425-022-2088-x

中国科学:数学(英文版)

中国科学院

中国科学:数学(英文版)

CSTPCD

影响因子：0.36

ISSN：1674-7283

年,卷(期)：2024.67(3)

参考文献量18