Robust PCA via tensor outlier pursuit

Jineng Ren; Xingguo Li; Jarvis D Haupt

doi:10.1109/ACSSC.2016.7869681

Robust PCA via tensor outlier pursuit

Jineng Ren, Xingguo Li, Jarvis D Haupt

Electrical and Computer Engineering

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

6 Scopus citations

Abstract

In this paper, we study robust principal component analysis on tensors, in the setting where frame-wise outliers exist. We propose a convex formulation to decompose a tensor into a low rank component and a frame-wise sparse component. Theoretically, we guarantee that exact subspace recovery and outlier identification can be achieved under mild model assumptions. Compared with entry-wise outlier pursuit and naive matricization of tensors with frame-wise outliers, our approach can handle higher ranks and proportion of outliers. Extensive numerical evaluations are provided on both synthetic and real data to support our theory.

Original language	English (US)
Title of host publication	Conference Record of the 50th Asilomar Conference on Signals, Systems and Computers, ACSSC 2016
Publisher	IEEE Computer Society
Pages	1744-1749
Number of pages	6
ISBN (Electronic)	9781538639542
DOIs	https://doi.org/10.1109/ACSSC.2016.7869681
State	Published - Mar 1 2017
Event	50th Asilomar Conference on Signals, Systems and Computers, ACSSC 2016 - Pacific Grove, United States Duration: Nov 6 2016 → Nov 9 2016

Other

Other	50th Asilomar Conference on Signals, Systems and Computers, ACSSC 2016
Country/Territory	United States
City	Pacific Grove
Period	11/6/16 → 11/9/16

Publisher link

10.1109/ACSSC.2016.7869681

Find It

Find It at U of M Libraries

Cite this

Ren, J, Li, X & Haupt, JD 2017, Robust PCA via tensor outlier pursuit. in Conference Record of the 50th Asilomar Conference on Signals, Systems and Computers, ACSSC 2016., 7869681, IEEE Computer Society, pp. 1744-1749, 50th Asilomar Conference on Signals, Systems and Computers, ACSSC 2016, Pacific Grove, United States, 11/6/16. https://doi.org/10.1109/ACSSC.2016.7869681

@inproceedings{e9f60c5a07584f9996f9386de34c2f69,

title = "Robust PCA via tensor outlier pursuit",

abstract = "In this paper, we study robust principal component analysis on tensors, in the setting where frame-wise outliers exist. We propose a convex formulation to decompose a tensor into a low rank component and a frame-wise sparse component. Theoretically, we guarantee that exact subspace recovery and outlier identification can be achieved under mild model assumptions. Compared with entry-wise outlier pursuit and naive matricization of tensors with frame-wise outliers, our approach can handle higher ranks and proportion of outliers. Extensive numerical evaluations are provided on both synthetic and real data to support our theory.",

author = "Jineng Ren and Xingguo Li and Haupt, {Jarvis D}",

year = "2017",

month = mar,

day = "1",

doi = "10.1109/ACSSC.2016.7869681",

language = "English (US)",

pages = "1744--1749",

booktitle = "Conference Record of the 50th Asilomar Conference on Signals, Systems and Computers, ACSSC 2016",

publisher = "IEEE Computer Society",

note = "50th Asilomar Conference on Signals, Systems and Computers, ACSSC 2016 ; Conference date: 06-11-2016 Through 09-11-2016",

}

TY - GEN

T1 - Robust PCA via tensor outlier pursuit

AU - Ren, Jineng

AU - Li, Xingguo

AU - Haupt, Jarvis D

PY - 2017/3/1

Y1 - 2017/3/1

N2 - In this paper, we study robust principal component analysis on tensors, in the setting where frame-wise outliers exist. We propose a convex formulation to decompose a tensor into a low rank component and a frame-wise sparse component. Theoretically, we guarantee that exact subspace recovery and outlier identification can be achieved under mild model assumptions. Compared with entry-wise outlier pursuit and naive matricization of tensors with frame-wise outliers, our approach can handle higher ranks and proportion of outliers. Extensive numerical evaluations are provided on both synthetic and real data to support our theory.

AB - In this paper, we study robust principal component analysis on tensors, in the setting where frame-wise outliers exist. We propose a convex formulation to decompose a tensor into a low rank component and a frame-wise sparse component. Theoretically, we guarantee that exact subspace recovery and outlier identification can be achieved under mild model assumptions. Compared with entry-wise outlier pursuit and naive matricization of tensors with frame-wise outliers, our approach can handle higher ranks and proportion of outliers. Extensive numerical evaluations are provided on both synthetic and real data to support our theory.

UR - http://www.scopus.com/inward/record.url?scp=85016241770&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=85016241770&partnerID=8YFLogxK

U2 - 10.1109/ACSSC.2016.7869681

DO - 10.1109/ACSSC.2016.7869681

M3 - Conference contribution

AN - SCOPUS:85016241770

SP - 1744

EP - 1749

BT - Conference Record of the 50th Asilomar Conference on Signals, Systems and Computers, ACSSC 2016

PB - IEEE Computer Society

T2 - 50th Asilomar Conference on Signals, Systems and Computers, ACSSC 2016

Y2 - 6 November 2016 through 9 November 2016

ER -

Robust PCA via tensor outlier pursuit

Abstract

Other

Publisher link

Find It

Other files and links

Fingerprint

Cite this