A feature ranking strategy to facilitate multivariate signal classification

Lalit Gupta, Srinivas Kota, Swetha Murali, Dennis L. Molfese, Ravi Vaidyanathan

Research output: Contribution to journalArticle

13 Citations (Scopus)

Abstract

A strategy is introduced to rank and select principal component transform (PCT) and discrete cosine transform (DCT) transform coefficient features to overcome the curse of dimensionality frequently encountered in implementing multivariate signal classifiers due to small sample sizes. The criteria considered for ranking include the magnitude, variance, interclass separation, and classification accuracies of the individual features. The feature ranking and selection strategy is applied to overcome the dimensionality problem, which often plagues the implementation and evaluation of practical Gaussian signal classifiers. The applications of the resulting PCT- and DCT-Gaussian signal classification strategies are demonstrated by classifying single-channel tonguemovement ear-pressure signals and multichannel event-related potentials. Through these experiments, it is shown that the dimension of the feature space can be decreased quite significantly by means of the feature ranking and selection strategy. The ranking strategy not only facilitates overcoming the dimensionality curse for multivariate classifier implementation but also provides a means to further select, out of a rank-ordered set, a smaller set of features that give the best classification accuracies. Results show that the PCT- andDCT-Gaussian classifiers yield higher classification accuracies than those reported in previous classification studies on the same signal sets. Among the combinations of the two transforms and four feature selection criteria, the PCT-Gaussian classifiers using the maximum magnitude and maximum variance selection criteria gave the best classification accuracies across the two sets of classification experiments. Most noteworthy is the fact that the multivariate Gaussian signal classifiers developed in this paper can be implemented without having to collect a prohibitively large number of training signals simply to satisfy the dimensionality conditions. Consequently, the classification strategies can be beneficial for designing personalized human-machine interface signal classifiers for individuals from whom only a limited number of training signals can reliably be collected due to severe disabilities.

Original languageEnglish (US)
Article number5170007
Pages (from-to)98-108
Number of pages11
JournalIEEE Transactions on Systems, Man and Cybernetics Part C: Applications and Reviews
Volume40
Issue number1
DOIs
StatePublished - Jan 1 2010

Fingerprint

Classifiers
Discrete cosine transforms
Feature extraction
Experiments

Keywords

  • Curse of dimensionality
  • Discrete cosine transform (DCT)
  • Ear-pressure signal classification
  • Event-related potential (ERP) classification
  • Feature ranking
  • Human-machine interface (HMI)
  • Multivariate signal classification
  • Principal component transform (PCT)

ASJC Scopus subject areas

  • Control and Systems Engineering
  • Software
  • Information Systems
  • Human-Computer Interaction
  • Computer Science Applications
  • Electrical and Electronic Engineering

Cite this

A feature ranking strategy to facilitate multivariate signal classification. / Gupta, Lalit; Kota, Srinivas; Murali, Swetha; Molfese, Dennis L.; Vaidyanathan, Ravi.

In: IEEE Transactions on Systems, Man and Cybernetics Part C: Applications and Reviews, Vol. 40, No. 1, 5170007, 01.01.2010, p. 98-108.

Research output: Contribution to journalArticle

@article{536a24388bdb42c6b00d1e3927306826,
title = "A feature ranking strategy to facilitate multivariate signal classification",
abstract = "A strategy is introduced to rank and select principal component transform (PCT) and discrete cosine transform (DCT) transform coefficient features to overcome the curse of dimensionality frequently encountered in implementing multivariate signal classifiers due to small sample sizes. The criteria considered for ranking include the magnitude, variance, interclass separation, and classification accuracies of the individual features. The feature ranking and selection strategy is applied to overcome the dimensionality problem, which often plagues the implementation and evaluation of practical Gaussian signal classifiers. The applications of the resulting PCT- and DCT-Gaussian signal classification strategies are demonstrated by classifying single-channel tonguemovement ear-pressure signals and multichannel event-related potentials. Through these experiments, it is shown that the dimension of the feature space can be decreased quite significantly by means of the feature ranking and selection strategy. The ranking strategy not only facilitates overcoming the dimensionality curse for multivariate classifier implementation but also provides a means to further select, out of a rank-ordered set, a smaller set of features that give the best classification accuracies. Results show that the PCT- andDCT-Gaussian classifiers yield higher classification accuracies than those reported in previous classification studies on the same signal sets. Among the combinations of the two transforms and four feature selection criteria, the PCT-Gaussian classifiers using the maximum magnitude and maximum variance selection criteria gave the best classification accuracies across the two sets of classification experiments. Most noteworthy is the fact that the multivariate Gaussian signal classifiers developed in this paper can be implemented without having to collect a prohibitively large number of training signals simply to satisfy the dimensionality conditions. Consequently, the classification strategies can be beneficial for designing personalized human-machine interface signal classifiers for individuals from whom only a limited number of training signals can reliably be collected due to severe disabilities.",
keywords = "Curse of dimensionality, Discrete cosine transform (DCT), Ear-pressure signal classification, Event-related potential (ERP) classification, Feature ranking, Human-machine interface (HMI), Multivariate signal classification, Principal component transform (PCT)",
author = "Lalit Gupta and Srinivas Kota and Swetha Murali and Molfese, {Dennis L.} and Ravi Vaidyanathan",
year = "2010",
month = "1",
day = "1",
doi = "10.1109/TSMCC.2009.2024648",
language = "English (US)",
volume = "40",
pages = "98--108",
journal = "IEEE Transactions on Systems, Man and Cybernetics Part C: Applications and Reviews",
issn = "1094-6977",
publisher = "Institute of Electrical and Electronics Engineers Inc.",
number = "1",

}

TY - JOUR

T1 - A feature ranking strategy to facilitate multivariate signal classification

AU - Gupta, Lalit

AU - Kota, Srinivas

AU - Murali, Swetha

AU - Molfese, Dennis L.

AU - Vaidyanathan, Ravi

PY - 2010/1/1

Y1 - 2010/1/1

N2 - A strategy is introduced to rank and select principal component transform (PCT) and discrete cosine transform (DCT) transform coefficient features to overcome the curse of dimensionality frequently encountered in implementing multivariate signal classifiers due to small sample sizes. The criteria considered for ranking include the magnitude, variance, interclass separation, and classification accuracies of the individual features. The feature ranking and selection strategy is applied to overcome the dimensionality problem, which often plagues the implementation and evaluation of practical Gaussian signal classifiers. The applications of the resulting PCT- and DCT-Gaussian signal classification strategies are demonstrated by classifying single-channel tonguemovement ear-pressure signals and multichannel event-related potentials. Through these experiments, it is shown that the dimension of the feature space can be decreased quite significantly by means of the feature ranking and selection strategy. The ranking strategy not only facilitates overcoming the dimensionality curse for multivariate classifier implementation but also provides a means to further select, out of a rank-ordered set, a smaller set of features that give the best classification accuracies. Results show that the PCT- andDCT-Gaussian classifiers yield higher classification accuracies than those reported in previous classification studies on the same signal sets. Among the combinations of the two transforms and four feature selection criteria, the PCT-Gaussian classifiers using the maximum magnitude and maximum variance selection criteria gave the best classification accuracies across the two sets of classification experiments. Most noteworthy is the fact that the multivariate Gaussian signal classifiers developed in this paper can be implemented without having to collect a prohibitively large number of training signals simply to satisfy the dimensionality conditions. Consequently, the classification strategies can be beneficial for designing personalized human-machine interface signal classifiers for individuals from whom only a limited number of training signals can reliably be collected due to severe disabilities.

AB - A strategy is introduced to rank and select principal component transform (PCT) and discrete cosine transform (DCT) transform coefficient features to overcome the curse of dimensionality frequently encountered in implementing multivariate signal classifiers due to small sample sizes. The criteria considered for ranking include the magnitude, variance, interclass separation, and classification accuracies of the individual features. The feature ranking and selection strategy is applied to overcome the dimensionality problem, which often plagues the implementation and evaluation of practical Gaussian signal classifiers. The applications of the resulting PCT- and DCT-Gaussian signal classification strategies are demonstrated by classifying single-channel tonguemovement ear-pressure signals and multichannel event-related potentials. Through these experiments, it is shown that the dimension of the feature space can be decreased quite significantly by means of the feature ranking and selection strategy. The ranking strategy not only facilitates overcoming the dimensionality curse for multivariate classifier implementation but also provides a means to further select, out of a rank-ordered set, a smaller set of features that give the best classification accuracies. Results show that the PCT- andDCT-Gaussian classifiers yield higher classification accuracies than those reported in previous classification studies on the same signal sets. Among the combinations of the two transforms and four feature selection criteria, the PCT-Gaussian classifiers using the maximum magnitude and maximum variance selection criteria gave the best classification accuracies across the two sets of classification experiments. Most noteworthy is the fact that the multivariate Gaussian signal classifiers developed in this paper can be implemented without having to collect a prohibitively large number of training signals simply to satisfy the dimensionality conditions. Consequently, the classification strategies can be beneficial for designing personalized human-machine interface signal classifiers for individuals from whom only a limited number of training signals can reliably be collected due to severe disabilities.

KW - Curse of dimensionality

KW - Discrete cosine transform (DCT)

KW - Ear-pressure signal classification

KW - Event-related potential (ERP) classification

KW - Feature ranking

KW - Human-machine interface (HMI)

KW - Multivariate signal classification

KW - Principal component transform (PCT)

UR - http://www.scopus.com/inward/record.url?scp=73049106805&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=73049106805&partnerID=8YFLogxK

U2 - 10.1109/TSMCC.2009.2024648

DO - 10.1109/TSMCC.2009.2024648

M3 - Article

AN - SCOPUS:73049106805

VL - 40

SP - 98

EP - 108

JO - IEEE Transactions on Systems, Man and Cybernetics Part C: Applications and Reviews

JF - IEEE Transactions on Systems, Man and Cybernetics Part C: Applications and Reviews

SN - 1094-6977

IS - 1

M1 - 5170007

ER -