The Linearized Alternating Direction Method of Multipliers for Dantzig Selector

Xiangfeng Wang; Xiaoming Yuan

doi:10.1137/110833543

The Linearized Alternating Direction Method of Multipliers for Dantzig Selector

Xiangfeng Wang, Xiaoming Yuan^*

^*Corresponding author for this work

Department of Mathematics

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

49 Citations (Scopus)

1 Downloads (Pure)

Abstract

The Dantzig selector was recently proposed to perform variable selection and model fitting in the linear regression model. It can be solved numerically by the alternating direction method of multipliers (ADM); and in this paper, we show that the application of ADM to the Dantzig selector can be speeded up significantly if one of its resulting subproblems at each iteration is linearized. The resulting linearized ADM for the Dantzig selector is shown to be efficient for solving both synthetic and real world data sets.

Original language	English
Pages (from-to)	A2792-A2811
Number of pages	20
Journal	SIAM Journal on Scientific Computing
Volume	34
Issue number	5
DOIs	https://doi.org/10.1137/110833543
Publication status	Published - 18 Oct 2012

Scopus Subject Areas

Computational Mathematics
Applied Mathematics

User-Defined Keywords

Alternating direction method of multipliers
Dantzig selector
Linear regression
Linearized

Access to Document

10.1137/110833543

Full TextFinal published version, 468 KB

Cite this

@article{58891ae01b894c7d867aa508338e509f,

title = "The Linearized Alternating Direction Method of Multipliers for Dantzig Selector",

abstract = "The Dantzig selector was recently proposed to perform variable selection and model fitting in the linear regression model. It can be solved numerically by the alternating direction method of multipliers (ADM); and in this paper, we show that the application of ADM to the Dantzig selector can be speeded up significantly if one of its resulting subproblems at each iteration is linearized. The resulting linearized ADM for the Dantzig selector is shown to be efficient for solving both synthetic and real world data sets.",

keywords = "Alternating direction method of multipliers, Dantzig selector, Linear regression, Linearized",

author = "Xiangfeng Wang and Xiaoming Yuan",

note = "Funding information: Department of Mathematics, Nanjing University, 22 Hankou Road, Nanjing, 210093, People{\textquoteright}s Republic of China (xfwang@nju.gmail.com). This author{\textquoteright}s work was supported by NSFC grant 91130007. ^ Corresponding author. Department of Mathematics, Hong Kong Baptist University, Kowloon Tong, Hong Kong, People{\textquoteright}s Republic of China (xmyuan@hkbu.edu.hk). This author{\textquoteright}s work was supported by the Hong Kong General Research Fund, 203311. Publisher copyright: {\textcopyright} 2012, Society for Industrial and Applied Mathematics",

year = "2012",

month = oct,

day = "18",

doi = "10.1137/110833543",

language = "English",

volume = "34",

pages = "A2792--A2811",

journal = "SIAM Journal of Scientific Computing",

issn = "1064-8275",

publisher = "Society for Industrial and Applied Mathematics (SIAM)",

number = "5",

}

TY - JOUR

T1 - The Linearized Alternating Direction Method of Multipliers for Dantzig Selector

AU - Wang, Xiangfeng

AU - Yuan, Xiaoming

N1 - Funding information: Department of Mathematics, Nanjing University, 22 Hankou Road, Nanjing, 210093, People’s Republic of China (xfwang@nju.gmail.com). This author’s work was supported by NSFC grant 91130007. ^ Corresponding author. Department of Mathematics, Hong Kong Baptist University, Kowloon Tong, Hong Kong, People’s Republic of China (xmyuan@hkbu.edu.hk). This author’s work was supported by the Hong Kong General Research Fund, 203311. Publisher copyright: © 2012, Society for Industrial and Applied Mathematics

PY - 2012/10/18

Y1 - 2012/10/18

N2 - The Dantzig selector was recently proposed to perform variable selection and model fitting in the linear regression model. It can be solved numerically by the alternating direction method of multipliers (ADM); and in this paper, we show that the application of ADM to the Dantzig selector can be speeded up significantly if one of its resulting subproblems at each iteration is linearized. The resulting linearized ADM for the Dantzig selector is shown to be efficient for solving both synthetic and real world data sets.

AB - The Dantzig selector was recently proposed to perform variable selection and model fitting in the linear regression model. It can be solved numerically by the alternating direction method of multipliers (ADM); and in this paper, we show that the application of ADM to the Dantzig selector can be speeded up significantly if one of its resulting subproblems at each iteration is linearized. The resulting linearized ADM for the Dantzig selector is shown to be efficient for solving both synthetic and real world data sets.

KW - Alternating direction method of multipliers

KW - Dantzig selector

KW - Linear regression

KW - Linearized

UR - http://www.scopus.com/inward/record.url?scp=84869774438&partnerID=8YFLogxK

U2 - 10.1137/110833543

DO - 10.1137/110833543

M3 - Article

AN - SCOPUS:84869774438

VL - 34

SP - A2792-A2811

JO - SIAM Journal of Scientific Computing

JF - SIAM Journal of Scientific Computing

SN - 1064-8275

IS - 5

ER -

The Linearized Alternating Direction Method of Multipliers for Dantzig Selector

Abstract

Scopus Subject Areas

User-Defined Keywords

Access to Document

Other files and links

Fingerprint

Cite this