RAP: Scalable RPCA for low-rank matrix recovery

Chong Peng; Zhao Kang; Ming Yang; Qiang Cheng

doi:10.1145/2983323.2983651

RAP: Scalable RPCA for low-rank matrix recovery

Chong Peng
, Zhao Kang
, Ming Yang
, Qiang Cheng

Internal Medicine

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

3 Scopus citations

Abstract

Recovering low-rank matrices is a problem common in many applications of data mining and machine learning, such as matrix completion and image denoising. Robust Principal Component Analysis (RPCA) has emerged for handling such kinds of problems; however, the existing RPCA approaches are usually computationally expensive, due to the fact that they need to obtain the singular value decomposition (SVD) of large matrices. In this paper, we propose a novel RPCA approach that eliminates the need for SVD of large matrices. Scalable algorithms are designed for several variants of our approach, which are crucial for real world applications on large scale data. Extensive experimental results confirm the effectiveness of our approach both quantitatively and visually.

Original language	English
Title of host publication	CIKM 2016 - Proceedings of the 2016 ACM Conference on Information and Knowledge Management
Pages	2113-2118
Number of pages	6
ISBN (Electronic)	9781450340731
DOIs	https://doi.org/10.1145/2983323.2983651
State	Published - Oct 24 2016
Event	25th ACM International Conference on Information and Knowledge Management, CIKM 2016 - Indianapolis, United States Duration: Oct 24 2016 → Oct 28 2016

Publication series

Name	International Conference on Information and Knowledge Management, Proceedings
Volume	24-28-October-2016

Conference

Conference	25th ACM International Conference on Information and Knowledge Management, CIKM 2016
Country/Territory	United States
City	Indianapolis
Period	10/24/16 → 10/28/16

Bibliographical note

Publisher Copyright:
© 2016 Copyright held by the owner/author(s).

Funding

This work is supported by National Science Foundation under grant IIS-1218712, National Natural Science Foundation of China, under grant 11241005, and Shanxi Scholarship Council of China 2015-093, Fund Program for the Scientific Activities of Selected Returned Overseas Professionals in Shanxi Province.

Funders	Funder number
Selected Returned Overseas Professionals in Shanxi Province
National Science Foundation Arctic Social Science Program	IIS-1218712
National Science Foundation Arctic Social Science Program
National Natural Science Foundation of China (NSFC)	11241005
National Natural Science Foundation of China (NSFC)
Shanxi Scholarship Council of China	2015-093
Shanxi Scholarship Council of China

Keywords

Fixed rank
Low-rank recovery
RPCA
Scabality

ASJC Scopus subject areas

General Decision Sciences
General Business, Management and Accounting

Access to Document

10.1145/2983323.2983651

Cite this

@inproceedings{eb82688d8d1f47f4b873d54b8b7bc181,

title = "RAP: Scalable RPCA for low-rank matrix recovery",

abstract = "Recovering low-rank matrices is a problem common in many applications of data mining and machine learning, such as matrix completion and image denoising. Robust Principal Component Analysis (RPCA) has emerged for handling such kinds of problems; however, the existing RPCA approaches are usually computationally expensive, due to the fact that they need to obtain the singular value decomposition (SVD) of large matrices. In this paper, we propose a novel RPCA approach that eliminates the need for SVD of large matrices. Scalable algorithms are designed for several variants of our approach, which are crucial for real world applications on large scale data. Extensive experimental results confirm the effectiveness of our approach both quantitatively and visually.",

keywords = "Fixed rank, Low-rank recovery, RPCA, Scabality",

author = "Chong Peng and Zhao Kang and Ming Yang and Qiang Cheng",

note = "Publisher Copyright: {\textcopyright} 2016 Copyright held by the owner/author(s).; 25th ACM International Conference on Information and Knowledge Management, CIKM 2016 ; Conference date: 24-10-2016 Through 28-10-2016",

year = "2016",

month = oct,

day = "24",

doi = "10.1145/2983323.2983651",

language = "English",

series = "International Conference on Information and Knowledge Management, Proceedings",

pages = "2113--2118",

booktitle = "CIKM 2016 - Proceedings of the 2016 ACM Conference on Information and Knowledge Management",

}

Peng, C, Kang, Z, Yang, M & Cheng, Q 2016, RAP: Scalable RPCA for low-rank matrix recovery. in CIKM 2016 - Proceedings of the 2016 ACM Conference on Information and Knowledge Management. International Conference on Information and Knowledge Management, Proceedings, vol. 24-28-October-2016, pp. 2113-2118, 25th ACM International Conference on Information and Knowledge Management, CIKM 2016, Indianapolis, United States, 10/24/16. https://doi.org/10.1145/2983323.2983651

RAP: Scalable RPCA for low-rank matrix recovery. / Peng, Chong; Kang, Zhao; Yang, Ming et al.
CIKM 2016 - Proceedings of the 2016 ACM Conference on Information and Knowledge Management. 2016. p. 2113-2118 (International Conference on Information and Knowledge Management, Proceedings; Vol. 24-28-October-2016).

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

TY - GEN

T1 - RAP

T2 - 25th ACM International Conference on Information and Knowledge Management, CIKM 2016

AU - Peng, Chong

AU - Kang, Zhao

AU - Yang, Ming

AU - Cheng, Qiang

PY - 2016/10/24

Y1 - 2016/10/24

N2 - Recovering low-rank matrices is a problem common in many applications of data mining and machine learning, such as matrix completion and image denoising. Robust Principal Component Analysis (RPCA) has emerged for handling such kinds of problems; however, the existing RPCA approaches are usually computationally expensive, due to the fact that they need to obtain the singular value decomposition (SVD) of large matrices. In this paper, we propose a novel RPCA approach that eliminates the need for SVD of large matrices. Scalable algorithms are designed for several variants of our approach, which are crucial for real world applications on large scale data. Extensive experimental results confirm the effectiveness of our approach both quantitatively and visually.

AB - Recovering low-rank matrices is a problem common in many applications of data mining and machine learning, such as matrix completion and image denoising. Robust Principal Component Analysis (RPCA) has emerged for handling such kinds of problems; however, the existing RPCA approaches are usually computationally expensive, due to the fact that they need to obtain the singular value decomposition (SVD) of large matrices. In this paper, we propose a novel RPCA approach that eliminates the need for SVD of large matrices. Scalable algorithms are designed for several variants of our approach, which are crucial for real world applications on large scale data. Extensive experimental results confirm the effectiveness of our approach both quantitatively and visually.

KW - Fixed rank

KW - Low-rank recovery

KW - RPCA

KW - Scabality

UR - https://www.scopus.com/pages/publications/84996490221

UR - https://www.scopus.com/pages/publications/84996490221#tab=citedBy

U2 - 10.1145/2983323.2983651

DO - 10.1145/2983323.2983651

M3 - Conference contribution

AN - SCOPUS:84996490221

T3 - International Conference on Information and Knowledge Management, Proceedings

SP - 2113

EP - 2118

BT - CIKM 2016 - Proceedings of the 2016 ACM Conference on Information and Knowledge Management

Y2 - 24 October 2016 through 28 October 2016

ER -

RAP: Scalable RPCA for low-rank matrix recovery

Abstract

Publication series

Conference

Bibliographical note

Funding

Keywords

ASJC Scopus subject areas

Access to Document

Other files and links

Fingerprint

Cite this