Enhancing the performance of feature selection algorithms for classifying hyperspectral imagery

Mukesh Kumar, Christopher J. Duffy, Patrick M. Reed

Research output: Contribution to conferencePaper

5 Citations (Scopus)

Abstract

A method for enhancing the performance of feature selection algorithms is proposed. The proposed method is a two step process - first a feature subset is selected with optimum mutual information content and then this subset is searched to find a smaller subset, which has the best separability between classes. A subset with "optimum" mutual information content is the one which contains most of the information that is present in the rest of set. An expression has been derived to find such a subset efficiently. The two-step process is shown to reduce the search space drastically. The method is implemented with a simple Genetic Algorithm (SGA) and tested using hyperspectral remote-sensing images (acquired by AVIRIS sensor) as a data set. Theoretical result shows that the proposed method reduces the computation load by 90%. A computational efficiency to the order ̃20% is obtained on the implementation of proposed method with SGA. The method is sufficiently general to be used to enhance other feature selection algorithms.

Original languageEnglish (US)
Pages3264-3267
Number of pages4
StatePublished - Dec 1 2004
Event2004 IEEE International Geoscience and Remote Sensing Symposium Proceedings: Science for Society: Exploring and Managing a Changing Planet. IGARSS 2004 - Anchorage, AK, United States
Duration: Sep 20 2004Sep 24 2004

Other

Other2004 IEEE International Geoscience and Remote Sensing Symposium Proceedings: Science for Society: Exploring and Managing a Changing Planet. IGARSS 2004
CountryUnited States
CityAnchorage, AK
Period9/20/049/24/04

Fingerprint

Feature extraction
imagery
Genetic algorithms
Computational efficiency
Remote sensing
genetic algorithm
Sensors
AVIRIS
method
sensor
remote sensing

All Science Journal Classification (ASJC) codes

  • Computer Science Applications
  • Earth and Planetary Sciences(all)

Cite this

Kumar, M., Duffy, C. J., & Reed, P. M. (2004). Enhancing the performance of feature selection algorithms for classifying hyperspectral imagery. 3264-3267. Paper presented at 2004 IEEE International Geoscience and Remote Sensing Symposium Proceedings: Science for Society: Exploring and Managing a Changing Planet. IGARSS 2004, Anchorage, AK, United States.
Kumar, Mukesh ; Duffy, Christopher J. ; Reed, Patrick M. / Enhancing the performance of feature selection algorithms for classifying hyperspectral imagery. Paper presented at 2004 IEEE International Geoscience and Remote Sensing Symposium Proceedings: Science for Society: Exploring and Managing a Changing Planet. IGARSS 2004, Anchorage, AK, United States.4 p.
@conference{df602b90b38e452fa02c89df7035aeb1,
title = "Enhancing the performance of feature selection algorithms for classifying hyperspectral imagery",
abstract = "A method for enhancing the performance of feature selection algorithms is proposed. The proposed method is a two step process - first a feature subset is selected with optimum mutual information content and then this subset is searched to find a smaller subset, which has the best separability between classes. A subset with {"}optimum{"} mutual information content is the one which contains most of the information that is present in the rest of set. An expression has been derived to find such a subset efficiently. The two-step process is shown to reduce the search space drastically. The method is implemented with a simple Genetic Algorithm (SGA) and tested using hyperspectral remote-sensing images (acquired by AVIRIS sensor) as a data set. Theoretical result shows that the proposed method reduces the computation load by 90{\%}. A computational efficiency to the order ̃20{\%} is obtained on the implementation of proposed method with SGA. The method is sufficiently general to be used to enhance other feature selection algorithms.",
author = "Mukesh Kumar and Duffy, {Christopher J.} and Reed, {Patrick M.}",
year = "2004",
month = "12",
day = "1",
language = "English (US)",
pages = "3264--3267",
note = "2004 IEEE International Geoscience and Remote Sensing Symposium Proceedings: Science for Society: Exploring and Managing a Changing Planet. IGARSS 2004 ; Conference date: 20-09-2004 Through 24-09-2004",

}

Kumar, M, Duffy, CJ & Reed, PM 2004, 'Enhancing the performance of feature selection algorithms for classifying hyperspectral imagery', Paper presented at 2004 IEEE International Geoscience and Remote Sensing Symposium Proceedings: Science for Society: Exploring and Managing a Changing Planet. IGARSS 2004, Anchorage, AK, United States, 9/20/04 - 9/24/04 pp. 3264-3267.

Enhancing the performance of feature selection algorithms for classifying hyperspectral imagery. / Kumar, Mukesh; Duffy, Christopher J.; Reed, Patrick M.

2004. 3264-3267 Paper presented at 2004 IEEE International Geoscience and Remote Sensing Symposium Proceedings: Science for Society: Exploring and Managing a Changing Planet. IGARSS 2004, Anchorage, AK, United States.

Research output: Contribution to conferencePaper

TY - CONF

T1 - Enhancing the performance of feature selection algorithms for classifying hyperspectral imagery

AU - Kumar, Mukesh

AU - Duffy, Christopher J.

AU - Reed, Patrick M.

PY - 2004/12/1

Y1 - 2004/12/1

N2 - A method for enhancing the performance of feature selection algorithms is proposed. The proposed method is a two step process - first a feature subset is selected with optimum mutual information content and then this subset is searched to find a smaller subset, which has the best separability between classes. A subset with "optimum" mutual information content is the one which contains most of the information that is present in the rest of set. An expression has been derived to find such a subset efficiently. The two-step process is shown to reduce the search space drastically. The method is implemented with a simple Genetic Algorithm (SGA) and tested using hyperspectral remote-sensing images (acquired by AVIRIS sensor) as a data set. Theoretical result shows that the proposed method reduces the computation load by 90%. A computational efficiency to the order ̃20% is obtained on the implementation of proposed method with SGA. The method is sufficiently general to be used to enhance other feature selection algorithms.

AB - A method for enhancing the performance of feature selection algorithms is proposed. The proposed method is a two step process - first a feature subset is selected with optimum mutual information content and then this subset is searched to find a smaller subset, which has the best separability between classes. A subset with "optimum" mutual information content is the one which contains most of the information that is present in the rest of set. An expression has been derived to find such a subset efficiently. The two-step process is shown to reduce the search space drastically. The method is implemented with a simple Genetic Algorithm (SGA) and tested using hyperspectral remote-sensing images (acquired by AVIRIS sensor) as a data set. Theoretical result shows that the proposed method reduces the computation load by 90%. A computational efficiency to the order ̃20% is obtained on the implementation of proposed method with SGA. The method is sufficiently general to be used to enhance other feature selection algorithms.

UR - http://www.scopus.com/inward/record.url?scp=15944423916&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=15944423916&partnerID=8YFLogxK

M3 - Paper

AN - SCOPUS:15944423916

SP - 3264

EP - 3267

ER -

Kumar M, Duffy CJ, Reed PM. Enhancing the performance of feature selection algorithms for classifying hyperspectral imagery. 2004. Paper presented at 2004 IEEE International Geoscience and Remote Sensing Symposium Proceedings: Science for Society: Exploring and Managing a Changing Planet. IGARSS 2004, Anchorage, AK, United States.