Robust principal component analysis based on low-rank and block-sparse matrix decomposition

Gongguo Tang; Arye Nehorai

doi:10.1109/CISS.2011.5766144

Robust principal component analysis based on low-rank and block-sparse matrix decomposition

Gongguo Tang, Arye Nehorai

Research output: Chapter in Book/Report/Conference proceeding › Conference contribution › peer-review

69 Scopus citations

Abstract

In this paper, we propose a convex program for low-rank and block-sparse matrix decomposition. Potential applications include outlier detection when certain columns of the data matrix are outliers. We design an algorithm based on the augmented Lagrange multiplier method to solve the convex program. We solve the subproblems involved in the augmented Lagrange multiplier method using the Douglas/Peaceman-Rachford (DR) monotone operator splitting method. Numerical simulations demonstrate the accuracy of our method compared with the robust principal component analysis based on low-rank and sparse matrix decomposition.

Original language	English
Title of host publication	2011 45th Annual Conference on Information Sciences and Systems, CISS 2011
DOIs	https://doi.org/10.1109/CISS.2011.5766144
State	Published - 2011
Event	2011 45th Annual Conference on Information Sciences and Systems, CISS 2011 - Baltimore, MD, United States Duration: Mar 23 2011 → Mar 25 2011

Publication series

Name	2011 45th Annual Conference on Information Sciences and Systems, CISS 2011

Conference

Conference	2011 45th Annual Conference on Information Sciences and Systems, CISS 2011
Country/Territory	United States
City	Baltimore, MD
Period	03/23/11 → 03/25/11

Keywords

augmented Lagrange multiplier method
low-rank and block-sparse matrix decomposition
operator splitting method
robust principal component analysis

Access to Document

10.1109/CISS.2011.5766144

Cite this

@inproceedings{305a612e1128460386e1b621e92f93bf,

title = "Robust principal component analysis based on low-rank and block-sparse matrix decomposition",

abstract = "In this paper, we propose a convex program for low-rank and block-sparse matrix decomposition. Potential applications include outlier detection when certain columns of the data matrix are outliers. We design an algorithm based on the augmented Lagrange multiplier method to solve the convex program. We solve the subproblems involved in the augmented Lagrange multiplier method using the Douglas/Peaceman-Rachford (DR) monotone operator splitting method. Numerical simulations demonstrate the accuracy of our method compared with the robust principal component analysis based on low-rank and sparse matrix decomposition.",

keywords = "augmented Lagrange multiplier method, low-rank and block-sparse matrix decomposition, operator splitting method, robust principal component analysis",

author = "Gongguo Tang and Arye Nehorai",

year = "2011",

doi = "10.1109/CISS.2011.5766144",

language = "English",

isbn = "9781424498475",

series = "2011 45th Annual Conference on Information Sciences and Systems, CISS 2011",

booktitle = "2011 45th Annual Conference on Information Sciences and Systems, CISS 2011",

note = "2011 45th Annual Conference on Information Sciences and Systems, CISS 2011 ; Conference date: 23-03-2011 Through 25-03-2011",

}

Tang, G & Nehorai, A 2011, Robust principal component analysis based on low-rank and block-sparse matrix decomposition. in 2011 45th Annual Conference on Information Sciences and Systems, CISS 2011., 5766144, 2011 45th Annual Conference on Information Sciences and Systems, CISS 2011, 2011 45th Annual Conference on Information Sciences and Systems, CISS 2011, Baltimore, MD, United States, 03/23/11. https://doi.org/10.1109/CISS.2011.5766144

Robust principal component analysis based on low-rank and block-sparse matrix decomposition. / Tang, Gongguo; Nehorai, Arye.
2011 45th Annual Conference on Information Sciences and Systems, CISS 2011. 2011. 5766144 (2011 45th Annual Conference on Information Sciences and Systems, CISS 2011).

Research output: Chapter in Book/Report/Conference proceeding › Conference contribution › peer-review

TY - GEN

T1 - Robust principal component analysis based on low-rank and block-sparse matrix decomposition

AU - Tang, Gongguo

AU - Nehorai, Arye

PY - 2011

Y1 - 2011

N2 - In this paper, we propose a convex program for low-rank and block-sparse matrix decomposition. Potential applications include outlier detection when certain columns of the data matrix are outliers. We design an algorithm based on the augmented Lagrange multiplier method to solve the convex program. We solve the subproblems involved in the augmented Lagrange multiplier method using the Douglas/Peaceman-Rachford (DR) monotone operator splitting method. Numerical simulations demonstrate the accuracy of our method compared with the robust principal component analysis based on low-rank and sparse matrix decomposition.

AB - In this paper, we propose a convex program for low-rank and block-sparse matrix decomposition. Potential applications include outlier detection when certain columns of the data matrix are outliers. We design an algorithm based on the augmented Lagrange multiplier method to solve the convex program. We solve the subproblems involved in the augmented Lagrange multiplier method using the Douglas/Peaceman-Rachford (DR) monotone operator splitting method. Numerical simulations demonstrate the accuracy of our method compared with the robust principal component analysis based on low-rank and sparse matrix decomposition.

KW - augmented Lagrange multiplier method

KW - low-rank and block-sparse matrix decomposition

KW - operator splitting method

KW - robust principal component analysis

UR - http://www.scopus.com/inward/record.url?scp=79957802748&partnerID=8YFLogxK

U2 - 10.1109/CISS.2011.5766144

DO - 10.1109/CISS.2011.5766144

M3 - Conference contribution

AN - SCOPUS:79957802748

SN - 9781424498475

T3 - 2011 45th Annual Conference on Information Sciences and Systems, CISS 2011

BT - 2011 45th Annual Conference on Information Sciences and Systems, CISS 2011

T2 - 2011 45th Annual Conference on Information Sciences and Systems, CISS 2011

Y2 - 23 March 2011 through 25 March 2011

ER -

Robust principal component analysis based on low-rank and block-sparse matrix decomposition

Abstract

Publication series

Conference

Keywords

Access to Document

Other files and links

Fingerprint

Cite this