Robust iteratively reweighted Lasso for sparse tensor factorizations

Hyon Jung Kim; Esa Ollila; Visa Koivunen; H. Vincent Poor

doi:10.1109/SSP.2014.6884665

Robust iteratively reweighted Lasso for sparse tensor factorizations

Hyon Jung Kim, Esa Ollila, Visa Koivunen, H. Vincent Poor

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

2 Scopus citations

Abstract

A new tensor approximation method is developed based on the CANDECOMP/PARAFAC (CP) factorization that enjoys both sparsity (i.e., yielding factor matrices with some non-zero elements) and resistance to outliers and non-Gaussian measurement noise. This method utilizes a robust bounded loss function for errors in the low-rank tensor approximation while encouraging sparsity with Lasso (or ℓ₁-) regularization to the factor matrices (of a tensor data). A simple alternating, iteratively reweighted (IRW) Lasso algorithm is proposed to solve the resulting optimization problem. Simulation studies illustrate that the proposed method provides excellent performance in terms of mean square error accuracy for heavy-tailed noise conditions, with relatively small loss in conventional Gaussian noise.

Original language	English (US)
Title of host publication	2014 IEEE Workshop on Statistical Signal Processing, SSP 2014
Publisher	IEEE Computer Society
Pages	420-423
Number of pages	4
ISBN (Print)	9781479949755
DOIs	https://doi.org/10.1109/SSP.2014.6884665
State	Published - 2014
Event	2014 IEEE Workshop on Statistical Signal Processing, SSP 2014 - Gold Coast, QLD, Australia Duration: Jun 29 2014 → Jul 2 2014

Publication series

Name	IEEE Workshop on Statistical Signal Processing Proceedings

Other

Other	2014 IEEE Workshop on Statistical Signal Processing, SSP 2014
Country	Australia
City	Gold Coast, QLD
Period	6/29/14 → 7/2/14

All Science Journal Classification (ASJC) codes

Electrical and Electronic Engineering
Applied Mathematics
Signal Processing
Computer Science Applications

Keywords

Iteratively reweighted least squares
Lasso
big data
regularization
robust loss function

Access to Document

10.1109/SSP.2014.6884665

Fingerprint Dive into the research topics of 'Robust iteratively reweighted Lasso for sparse tensor factorizations'. Together they form a unique fingerprint.

Cite this

@inproceedings{f21200e1ae7e461fa750546dc73139f6,

title = "Robust iteratively reweighted Lasso for sparse tensor factorizations",

abstract = "A new tensor approximation method is developed based on the CANDECOMP/PARAFAC (CP) factorization that enjoys both sparsity (i.e., yielding factor matrices with some non-zero elements) and resistance to outliers and non-Gaussian measurement noise. This method utilizes a robust bounded loss function for errors in the low-rank tensor approximation while encouraging sparsity with Lasso (or ℓ1-) regularization to the factor matrices (of a tensor data). A simple alternating, iteratively reweighted (IRW) Lasso algorithm is proposed to solve the resulting optimization problem. Simulation studies illustrate that the proposed method provides excellent performance in terms of mean square error accuracy for heavy-tailed noise conditions, with relatively small loss in conventional Gaussian noise.",

keywords = "Iteratively reweighted least squares, Lasso, big data, regularization, robust loss function",

author = "Kim, {Hyon Jung} and Esa Ollila and Visa Koivunen and Poor, {H. Vincent}",

year = "2014",

doi = "10.1109/SSP.2014.6884665",

language = "English (US)",

isbn = "9781479949755",

series = "IEEE Workshop on Statistical Signal Processing Proceedings",

publisher = "IEEE Computer Society",

pages = "420--423",

booktitle = "2014 IEEE Workshop on Statistical Signal Processing, SSP 2014",

address = "United States",

}

Kim, HJ, Ollila, E, Koivunen, V & Poor, HV 2014, Robust iteratively reweighted Lasso for sparse tensor factorizations. in 2014 IEEE Workshop on Statistical Signal Processing, SSP 2014., 6884665, IEEE Workshop on Statistical Signal Processing Proceedings, IEEE Computer Society, pp. 420-423, 2014 IEEE Workshop on Statistical Signal Processing, SSP 2014, Gold Coast, QLD, Australia, 6/29/14. https://doi.org/10.1109/SSP.2014.6884665

Robust iteratively reweighted Lasso for sparse tensor factorizations. / Kim, Hyon Jung; Ollila, Esa; Koivunen, Visa; Poor, H. Vincent.

2014 IEEE Workshop on Statistical Signal Processing, SSP 2014. IEEE Computer Society, 2014. p. 420-423 6884665 (IEEE Workshop on Statistical Signal Processing Proceedings).

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

TY - GEN

T1 - Robust iteratively reweighted Lasso for sparse tensor factorizations

AU - Kim, Hyon Jung

AU - Ollila, Esa

AU - Koivunen, Visa

AU - Poor, H. Vincent

PY - 2014

Y1 - 2014

N2 - A new tensor approximation method is developed based on the CANDECOMP/PARAFAC (CP) factorization that enjoys both sparsity (i.e., yielding factor matrices with some non-zero elements) and resistance to outliers and non-Gaussian measurement noise. This method utilizes a robust bounded loss function for errors in the low-rank tensor approximation while encouraging sparsity with Lasso (or ℓ1-) regularization to the factor matrices (of a tensor data). A simple alternating, iteratively reweighted (IRW) Lasso algorithm is proposed to solve the resulting optimization problem. Simulation studies illustrate that the proposed method provides excellent performance in terms of mean square error accuracy for heavy-tailed noise conditions, with relatively small loss in conventional Gaussian noise.

AB - A new tensor approximation method is developed based on the CANDECOMP/PARAFAC (CP) factorization that enjoys both sparsity (i.e., yielding factor matrices with some non-zero elements) and resistance to outliers and non-Gaussian measurement noise. This method utilizes a robust bounded loss function for errors in the low-rank tensor approximation while encouraging sparsity with Lasso (or ℓ1-) regularization to the factor matrices (of a tensor data). A simple alternating, iteratively reweighted (IRW) Lasso algorithm is proposed to solve the resulting optimization problem. Simulation studies illustrate that the proposed method provides excellent performance in terms of mean square error accuracy for heavy-tailed noise conditions, with relatively small loss in conventional Gaussian noise.

KW - Iteratively reweighted least squares

KW - Lasso

KW - big data

KW - regularization

KW - robust loss function

UR - http://www.scopus.com/inward/record.url?scp=84907391371&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=84907391371&partnerID=8YFLogxK

U2 - 10.1109/SSP.2014.6884665

DO - 10.1109/SSP.2014.6884665

M3 - Conference contribution

AN - SCOPUS:84907391371

SN - 9781479949755

T3 - IEEE Workshop on Statistical Signal Processing Proceedings

SP - 420

EP - 423

BT - 2014 IEEE Workshop on Statistical Signal Processing, SSP 2014

PB - IEEE Computer Society

T2 - 2014 IEEE Workshop on Statistical Signal Processing, SSP 2014

Y2 - 29 June 2014 through 2 July 2014

ER -

Robust iteratively reweighted Lasso for sparse tensor factorizations

Abstract

Publication series

Other

All Science Journal Classification (ASJC) codes

Keywords

Access to Document

Other files and links

Cite this