A note on shrinkage sliced inverse regression

Liqiang Ni; R. Dennis Cook; Chih Ling Tsai

doi:10.1093/biomet/92.1.242

A note on shrinkage sliced inverse regression

Liqiang Ni, R. Dennis Cook, Chih Ling Tsai

Statistics (Twin Cities)

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

66 Scopus citations

Abstract

We employ Lasso shrinkage within the context of sufficient dimension reduction to obtain a shrinkage sliced inverse regression estimator, which provides easier interpretations and better prediction accuracy without assuming a parametric model. The shrinkage sliced inverse regression approach can be employed for both single-index and multiple-index models. Simulation studies suggest that the new estimator performs well when its tuning parameter is selected by either the Bayesian information criterion or the residual information criterion.

Original language	English (US)
Pages (from-to)	242-247
Number of pages	6
Journal	Biometrika
Volume	92
Issue number	1
DOIs	https://doi.org/10.1093/biomet/92.1.242
State	Published - Mar 2005

Bibliographical note

Funding Information:
We thank the referee and editor for their useful comments. R. D. Cook’s research was supported in part by the U.S. National Science Foundation and Chih-Ling Tsai’s research was supported in part by the U.S. National Institutes of Health.

Keywords

Garotte
Lasso
Shrinkage estimator
Sliced inverse regression
Sufficient dimension reduction

Publisher link

10.1093/biomet/92.1.242

Find It

Find It at U of M Libraries

Cite this

@article{721415ba11414f859d85da209f41b56c,

title = "A note on shrinkage sliced inverse regression",

abstract = "We employ Lasso shrinkage within the context of sufficient dimension reduction to obtain a shrinkage sliced inverse regression estimator, which provides easier interpretations and better prediction accuracy without assuming a parametric model. The shrinkage sliced inverse regression approach can be employed for both single-index and multiple-index models. Simulation studies suggest that the new estimator performs well when its tuning parameter is selected by either the Bayesian information criterion or the residual information criterion.",

keywords = "Garotte, Lasso, Shrinkage estimator, Sliced inverse regression, Sufficient dimension reduction",

author = "Liqiang Ni and Cook, \{R. Dennis\} and Tsai, \{Chih Ling\}",

note = "Funding Information: We thank the referee and editor for their useful comments. R. D. Cook{\textquoteright}s research was supported in part by the U.S. National Science Foundation and Chih-Ling Tsai{\textquoteright}s research was supported in part by the U.S. National Institutes of Health.",

year = "2005",

month = mar,

doi = "10.1093/biomet/92.1.242",

language = "English (US)",

volume = "92",

pages = "242--247",

journal = "Biometrika",

issn = "0006-3444",

publisher = "Oxford University Press",

number = "1",

}

TY - JOUR

T1 - A note on shrinkage sliced inverse regression

AU - Ni, Liqiang

AU - Cook, R. Dennis

AU - Tsai, Chih Ling

N1 - Funding Information: We thank the referee and editor for their useful comments. R. D. Cook’s research was supported in part by the U.S. National Science Foundation and Chih-Ling Tsai’s research was supported in part by the U.S. National Institutes of Health.

PY - 2005/3

Y1 - 2005/3

N2 - We employ Lasso shrinkage within the context of sufficient dimension reduction to obtain a shrinkage sliced inverse regression estimator, which provides easier interpretations and better prediction accuracy without assuming a parametric model. The shrinkage sliced inverse regression approach can be employed for both single-index and multiple-index models. Simulation studies suggest that the new estimator performs well when its tuning parameter is selected by either the Bayesian information criterion or the residual information criterion.

AB - We employ Lasso shrinkage within the context of sufficient dimension reduction to obtain a shrinkage sliced inverse regression estimator, which provides easier interpretations and better prediction accuracy without assuming a parametric model. The shrinkage sliced inverse regression approach can be employed for both single-index and multiple-index models. Simulation studies suggest that the new estimator performs well when its tuning parameter is selected by either the Bayesian information criterion or the residual information criterion.

KW - Garotte

KW - Lasso

KW - Shrinkage estimator

KW - Sliced inverse regression

KW - Sufficient dimension reduction

UR - https://www.scopus.com/pages/publications/15844375310

UR - https://www.scopus.com/inward/citedby.url?scp=15844375310&partnerID=8YFLogxK

U2 - 10.1093/biomet/92.1.242

DO - 10.1093/biomet/92.1.242

M3 - Article

AN - SCOPUS:15844375310

SN - 0006-3444

VL - 92

SP - 242

EP - 247

JO - Biometrika

JF - Biometrika

IS - 1

ER -

A note on shrinkage sliced inverse regression

Abstract

Bibliographical note

Keywords

Publisher link

Find It

Other files and links

Fingerprint

Cite this