Sparse sensor selection for nonparametric decentralized detection via L1 regularization

Weiguang Wang, Yingbin Liang, Eric P. Xing, Lixin Shen

Research output: Chapter in Book/Report/Conference proceedingConference contribution

Abstract

Sensor selection in nonparametric decentralized detection is investigated. Kernel-based minimization framework with a weighted kernel is adopted, where the kernel weight parameters represent sensors' contributions to decision making. L1 regularization on weight parameters is introduced into the risk function so that the resulting optimal decision rule contains a sparse vector of nonzero weight parameters. In this way, sensor selection is naturally performed because only sensors corresponding to nonzero weight parameters contribute to decision making. A gradient projection algorithm and a Gauss-Seidel algorithm are developed to jointly perform weight selection (i.e., sensor selection) and optimize decision rules. Both algorithms are shown to converge to critical points for this non-convex optimization problem. Numerical results are provided to demonstrate the advantages and properties of the proposed sensor selection approach.

Original languageEnglish (US)
Title of host publicationIEEE International Workshop on Machine Learning for Signal Processing, MLSP
PublisherIEEE Computer Society
ISBN (Print)9781479936946
DOIs
StatePublished - Nov 14 2014
Event2014 24th IEEE International Workshop on Machine Learning for Signal Processing, MLSP 2014 - Reims, France
Duration: Sep 21 2014Sep 24 2014

Other

Other2014 24th IEEE International Workshop on Machine Learning for Signal Processing, MLSP 2014
CountryFrance
CityReims
Period9/21/149/24/14

Fingerprint

Sensors
Decision making

ASJC Scopus subject areas

  • Human-Computer Interaction
  • Signal Processing

Cite this

Wang, W., Liang, Y., Xing, E. P., & Shen, L. (2014). Sparse sensor selection for nonparametric decentralized detection via L1 regularization. In IEEE International Workshop on Machine Learning for Signal Processing, MLSP [6958898] IEEE Computer Society. https://doi.org/10.1109/MLSP.2014.6958898

Sparse sensor selection for nonparametric decentralized detection via L1 regularization. / Wang, Weiguang; Liang, Yingbin; Xing, Eric P.; Shen, Lixin.

IEEE International Workshop on Machine Learning for Signal Processing, MLSP. IEEE Computer Society, 2014. 6958898.

Research output: Chapter in Book/Report/Conference proceedingConference contribution

Wang, W, Liang, Y, Xing, EP & Shen, L 2014, Sparse sensor selection for nonparametric decentralized detection via L1 regularization. in IEEE International Workshop on Machine Learning for Signal Processing, MLSP., 6958898, IEEE Computer Society, 2014 24th IEEE International Workshop on Machine Learning for Signal Processing, MLSP 2014, Reims, France, 9/21/14. https://doi.org/10.1109/MLSP.2014.6958898
Wang W, Liang Y, Xing EP, Shen L. Sparse sensor selection for nonparametric decentralized detection via L1 regularization. In IEEE International Workshop on Machine Learning for Signal Processing, MLSP. IEEE Computer Society. 2014. 6958898 https://doi.org/10.1109/MLSP.2014.6958898
Wang, Weiguang ; Liang, Yingbin ; Xing, Eric P. ; Shen, Lixin. / Sparse sensor selection for nonparametric decentralized detection via L1 regularization. IEEE International Workshop on Machine Learning for Signal Processing, MLSP. IEEE Computer Society, 2014.
@inproceedings{31d84e5aa72a435cb0c56af39fee75f0,
title = "Sparse sensor selection for nonparametric decentralized detection via L1 regularization",
abstract = "Sensor selection in nonparametric decentralized detection is investigated. Kernel-based minimization framework with a weighted kernel is adopted, where the kernel weight parameters represent sensors' contributions to decision making. L1 regularization on weight parameters is introduced into the risk function so that the resulting optimal decision rule contains a sparse vector of nonzero weight parameters. In this way, sensor selection is naturally performed because only sensors corresponding to nonzero weight parameters contribute to decision making. A gradient projection algorithm and a Gauss-Seidel algorithm are developed to jointly perform weight selection (i.e., sensor selection) and optimize decision rules. Both algorithms are shown to converge to critical points for this non-convex optimization problem. Numerical results are provided to demonstrate the advantages and properties of the proposed sensor selection approach.",
author = "Weiguang Wang and Yingbin Liang and Xing, {Eric P.} and Lixin Shen",
year = "2014",
month = "11",
day = "14",
doi = "10.1109/MLSP.2014.6958898",
language = "English (US)",
isbn = "9781479936946",
booktitle = "IEEE International Workshop on Machine Learning for Signal Processing, MLSP",
publisher = "IEEE Computer Society",
address = "United States",

}

TY - GEN

T1 - Sparse sensor selection for nonparametric decentralized detection via L1 regularization

AU - Wang, Weiguang

AU - Liang, Yingbin

AU - Xing, Eric P.

AU - Shen, Lixin

PY - 2014/11/14

Y1 - 2014/11/14

N2 - Sensor selection in nonparametric decentralized detection is investigated. Kernel-based minimization framework with a weighted kernel is adopted, where the kernel weight parameters represent sensors' contributions to decision making. L1 regularization on weight parameters is introduced into the risk function so that the resulting optimal decision rule contains a sparse vector of nonzero weight parameters. In this way, sensor selection is naturally performed because only sensors corresponding to nonzero weight parameters contribute to decision making. A gradient projection algorithm and a Gauss-Seidel algorithm are developed to jointly perform weight selection (i.e., sensor selection) and optimize decision rules. Both algorithms are shown to converge to critical points for this non-convex optimization problem. Numerical results are provided to demonstrate the advantages and properties of the proposed sensor selection approach.

AB - Sensor selection in nonparametric decentralized detection is investigated. Kernel-based minimization framework with a weighted kernel is adopted, where the kernel weight parameters represent sensors' contributions to decision making. L1 regularization on weight parameters is introduced into the risk function so that the resulting optimal decision rule contains a sparse vector of nonzero weight parameters. In this way, sensor selection is naturally performed because only sensors corresponding to nonzero weight parameters contribute to decision making. A gradient projection algorithm and a Gauss-Seidel algorithm are developed to jointly perform weight selection (i.e., sensor selection) and optimize decision rules. Both algorithms are shown to converge to critical points for this non-convex optimization problem. Numerical results are provided to demonstrate the advantages and properties of the proposed sensor selection approach.

UR - http://www.scopus.com/inward/record.url?scp=84912570248&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=84912570248&partnerID=8YFLogxK

U2 - 10.1109/MLSP.2014.6958898

DO - 10.1109/MLSP.2014.6958898

M3 - Conference contribution

SN - 9781479936946

BT - IEEE International Workshop on Machine Learning for Signal Processing, MLSP

PB - IEEE Computer Society

ER -