Dimension reduction via local rank regression

Yuexiao Dong; Bo Kai; Zhou Yu

doi:10.1080/00949655.2016.1205067

Dimension reduction via local rank regression

Yuexiao Dong
, Bo Kai
, Zhou Yu^*

^*Corresponding author for this work

School of Statistics

Research output: Contribution to journal › Article › peer-review

1 Scopus citations

Abstract

Outer product of gradients (OPG) achieves dimension reduction via estimating the gradients of the regression function. In this paper, we propose two novel OPG estimators via local rank regression: the rank OPG estimator and the Walsh-average OPG estimator. Both proposals guard against a wide range of error distributions, and are safe alternatives to existing OPG estimators based on local linear regression or local L1 regression. The effectiveness of the new proposals are demonstrated via extensive numerical studies.

Original language	English
Pages (from-to)	239-249
Number of pages	11
Journal	Journal of Statistical Computation and Simulation
Volume	87
Issue number	2
DOIs	https://doi.org/10.1080/00949655.2016.1205067
State	Published - 22 Jan 2017

Keywords

Local Walsh-average regression
non-parametric regression
order determination
outer product gradient

Access to Document

10.1080/00949655.2016.1205067

Cite this

@article{a4baebefa85444bb93a3bbef0e2fe4b3,

title = "Dimension reduction via local rank regression",

abstract = "Outer product of gradients (OPG) achieves dimension reduction via estimating the gradients of the regression function. In this paper, we propose two novel OPG estimators via local rank regression: the rank OPG estimator and the Walsh-average OPG estimator. Both proposals guard against a wide range of error distributions, and are safe alternatives to existing OPG estimators based on local linear regression or local L1 regression. The effectiveness of the new proposals are demonstrated via extensive numerical studies.",

keywords = "Local Walsh-average regression, non-parametric regression, order determination, outer product gradient",

author = "Yuexiao Dong and Bo Kai and Zhou Yu",

note = "Publisher Copyright: {\textcopyright} 2016 Informa UK Limited, trading as Taylor \& Francis Group.",

year = "2017",

month = jan,

day = "22",

doi = "10.1080/00949655.2016.1205067",

language = "英语",

volume = "87",

pages = "239--249",

journal = "Journal of Statistical Computation and Simulation",

issn = "0094-9655",

publisher = "Taylor and Francis Ltd.",

number = "2",

}

TY - JOUR

T1 - Dimension reduction via local rank regression

AU - Dong, Yuexiao

AU - Kai, Bo

AU - Yu, Zhou

PY - 2017/1/22

Y1 - 2017/1/22

N2 - Outer product of gradients (OPG) achieves dimension reduction via estimating the gradients of the regression function. In this paper, we propose two novel OPG estimators via local rank regression: the rank OPG estimator and the Walsh-average OPG estimator. Both proposals guard against a wide range of error distributions, and are safe alternatives to existing OPG estimators based on local linear regression or local L1 regression. The effectiveness of the new proposals are demonstrated via extensive numerical studies.

AB - Outer product of gradients (OPG) achieves dimension reduction via estimating the gradients of the regression function. In this paper, we propose two novel OPG estimators via local rank regression: the rank OPG estimator and the Walsh-average OPG estimator. Both proposals guard against a wide range of error distributions, and are safe alternatives to existing OPG estimators based on local linear regression or local L1 regression. The effectiveness of the new proposals are demonstrated via extensive numerical studies.

KW - Local Walsh-average regression

KW - non-parametric regression

KW - order determination

KW - outer product gradient

UR - https://www.scopus.com/pages/publications/84978499739

U2 - 10.1080/00949655.2016.1205067

DO - 10.1080/00949655.2016.1205067

M3 - 文章

AN - SCOPUS:84978499739

SN - 0094-9655

VL - 87

SP - 239

EP - 249

JO - Journal of Statistical Computation and Simulation

JF - Journal of Statistical Computation and Simulation

IS - 2

ER -

Dimension reduction via local rank regression

Abstract

Keywords

Access to Document

Other files and links

Fingerprint

Cite this