Sparsity-enforced regression based on over-complete dictionary

Peng Yang; Gongguo Tang; Arye Nehorai

doi:10.1109/CAMSAP.2011.6135999

Sparsity-enforced regression based on over-complete dictionary

Peng Yang, Gongguo Tang, Arye Nehorai

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

1 Scopus citations

Abstract

Nonlinear regression has broad applications in various research areas, and kernel-based regression is very popular in machine learning literature. However, the selection of basis-function parameters is often difficult. In this paper we propose a new sparsity-enforced regression method based on an over-complete dictionary. The over-complete dictionary comprises basis functions with quantized parameters, and we employ ₁-regularized minimization to obtain a sparse weight vector of the basis. The ₁-regularized minimization automatically selects the most suitable basis function parameters. Performance analysis shows that this new method provides improved regression accuracy with small model complexity as measured by the number of non-zero entries of the weight vector.

Original language	English
Title of host publication	2011 4th IEEE International Workshop on Computational Advances in Multi-Sensor Adaptive Processing, CAMSAP 2011
Pages	261-264
Number of pages	4
DOIs	https://doi.org/10.1109/CAMSAP.2011.6135999
State	Published - 2011
Event	2011 4th IEEE International Workshop on Computational Advances in Multi-Sensor Adaptive Processing, CAMSAP 2011 - San Juan, Puerto Rico Duration: Dec 13 2011 → Dec 16 2011

Publication series

Name	2011 4th IEEE International Workshop on Computational Advances in Multi-Sensor Adaptive Processing, CAMSAP 2011

Conference

Conference	2011 4th IEEE International Workshop on Computational Advances in Multi-Sensor Adaptive Processing, CAMSAP 2011
Country/Territory	Puerto Rico
City	San Juan
Period	12/13/11 → 12/16/11

Access to Document

10.1109/CAMSAP.2011.6135999

Cite this

Yang, P., Tang, G., & Nehorai, A. (2011). Sparsity-enforced regression based on over-complete dictionary. In 2011 4th IEEE International Workshop on Computational Advances in Multi-Sensor Adaptive Processing, CAMSAP 2011 (pp. 261-264). Article 6135999 (2011 4th IEEE International Workshop on Computational Advances in Multi-Sensor Adaptive Processing, CAMSAP 2011). https://doi.org/10.1109/CAMSAP.2011.6135999

@inproceedings{e2050b7a757841d49c5cc8b8a57ee492,

title = "Sparsity-enforced regression based on over-complete dictionary",

abstract = "Nonlinear regression has broad applications in various research areas, and kernel-based regression is very popular in machine learning literature. However, the selection of basis-function parameters is often difficult. In this paper we propose a new sparsity-enforced regression method based on an over-complete dictionary. The over-complete dictionary comprises basis functions with quantized parameters, and we employ 1-regularized minimization to obtain a sparse weight vector of the basis. The 1-regularized minimization automatically selects the most suitable basis function parameters. Performance analysis shows that this new method provides improved regression accuracy with small model complexity as measured by the number of non-zero entries of the weight vector.",

author = "Peng Yang and Gongguo Tang and Arye Nehorai",

year = "2011",

doi = "10.1109/CAMSAP.2011.6135999",

language = "English",

isbn = "9781457721052",

series = "2011 4th IEEE International Workshop on Computational Advances in Multi-Sensor Adaptive Processing, CAMSAP 2011",

pages = "261--264",

booktitle = "2011 4th IEEE International Workshop on Computational Advances in Multi-Sensor Adaptive Processing, CAMSAP 2011",

note = "2011 4th IEEE International Workshop on Computational Advances in Multi-Sensor Adaptive Processing, CAMSAP 2011 ; Conference date: 13-12-2011 Through 16-12-2011",

}

Yang, P, Tang, G & Nehorai, A 2011, Sparsity-enforced regression based on over-complete dictionary. in 2011 4th IEEE International Workshop on Computational Advances in Multi-Sensor Adaptive Processing, CAMSAP 2011., 6135999, 2011 4th IEEE International Workshop on Computational Advances in Multi-Sensor Adaptive Processing, CAMSAP 2011, pp. 261-264, 2011 4th IEEE International Workshop on Computational Advances in Multi-Sensor Adaptive Processing, CAMSAP 2011, San Juan, Puerto Rico, 12/13/11. https://doi.org/10.1109/CAMSAP.2011.6135999

Sparsity-enforced regression based on over-complete dictionary. / Yang, Peng; Tang, Gongguo; Nehorai, Arye.
2011 4th IEEE International Workshop on Computational Advances in Multi-Sensor Adaptive Processing, CAMSAP 2011. 2011. p. 261-264 6135999 (2011 4th IEEE International Workshop on Computational Advances in Multi-Sensor Adaptive Processing, CAMSAP 2011).

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

TY - GEN

T1 - Sparsity-enforced regression based on over-complete dictionary

AU - Yang, Peng

AU - Tang, Gongguo

AU - Nehorai, Arye

PY - 2011

Y1 - 2011

N2 - Nonlinear regression has broad applications in various research areas, and kernel-based regression is very popular in machine learning literature. However, the selection of basis-function parameters is often difficult. In this paper we propose a new sparsity-enforced regression method based on an over-complete dictionary. The over-complete dictionary comprises basis functions with quantized parameters, and we employ 1-regularized minimization to obtain a sparse weight vector of the basis. The 1-regularized minimization automatically selects the most suitable basis function parameters. Performance analysis shows that this new method provides improved regression accuracy with small model complexity as measured by the number of non-zero entries of the weight vector.

AB - Nonlinear regression has broad applications in various research areas, and kernel-based regression is very popular in machine learning literature. However, the selection of basis-function parameters is often difficult. In this paper we propose a new sparsity-enforced regression method based on an over-complete dictionary. The over-complete dictionary comprises basis functions with quantized parameters, and we employ 1-regularized minimization to obtain a sparse weight vector of the basis. The 1-regularized minimization automatically selects the most suitable basis function parameters. Performance analysis shows that this new method provides improved regression accuracy with small model complexity as measured by the number of non-zero entries of the weight vector.

UR - http://www.scopus.com/inward/record.url?scp=84857144461&partnerID=8YFLogxK

U2 - 10.1109/CAMSAP.2011.6135999

DO - 10.1109/CAMSAP.2011.6135999

M3 - Conference contribution

AN - SCOPUS:84857144461

SN - 9781457721052

T3 - 2011 4th IEEE International Workshop on Computational Advances in Multi-Sensor Adaptive Processing, CAMSAP 2011

SP - 261

EP - 264

BT - 2011 4th IEEE International Workshop on Computational Advances in Multi-Sensor Adaptive Processing, CAMSAP 2011

T2 - 2011 4th IEEE International Workshop on Computational Advances in Multi-Sensor Adaptive Processing, CAMSAP 2011

Y2 - 13 December 2011 through 16 December 2011

ER -

Yang P, Tang G, Nehorai A. Sparsity-enforced regression based on over-complete dictionary. In 2011 4th IEEE International Workshop on Computational Advances in Multi-Sensor Adaptive Processing, CAMSAP 2011. 2011. p. 261-264. 6135999. (2011 4th IEEE International Workshop on Computational Advances in Multi-Sensor Adaptive Processing, CAMSAP 2011). doi: 10.1109/CAMSAP.2011.6135999

Sparsity-enforced regression based on over-complete dictionary

Abstract

Publication series

Conference

Access to Document

Other files and links

Fingerprint

Cite this