Feature selection using misclassification counts

Adil Bagiriv; Andrew Yatsko; Andrew Stranieri; Herbert Jelinek

Feature selection using misclassification counts

Adil Bagiriv, Andrew Yatsko, Andrew Stranieri, Herbert Jelinek

University of Ballarat

Research output: Book chapter/Published conference paper › Conference paper › peer-review

Abstract

Dimensionality reduction of the problem space through detection and removal of variables, contributing little or not at all to classification, is able to relieve the computational load and instance acquisition effort, considering all the data attributes accessed each time around. The approach to feature selection in this paper is based on the concept of coherent accumulation of data about class centers with respect to coordinates of informative features. Ranking is done on the degree to which different variables exhibit random characteristics. The results are being verified using the Nearest Neighbor classifier. This also helps to address the feature irrelevance and redundancy, what ranking does not immediately decide. Additionally, feature ranking methods from different independent sources are called in for the direct comparison.

Original language	English
Title of host publication	Proceedings of the Ninth Australasian Data Mining Conference (AusDM 11)
Editors	P Kennedy P Kennedy
Place of Publication	Sydney, Australia
Publisher	Australian Computer Society Inc
Pages	51-62
Number of pages	12
Volume	121
Publication status	Published - 2011
Event	The 9th Australasian Data Mining Conference: AusDM 2011 - University of Ballarat, Ballarat, Australia Duration: 01 Dec 2011 → 02 Dec 2011 http://www.wikicfp.com/cfp/servlet/event.showcfp?eventid=16402&copyownerid=3591

Publication series

Name
ISSN (Print)	1445-1336

Conference

Conference	The 9th Australasian Data Mining Conference
Country/Territory	Australia
City	Ballarat
Period	01/12/11 → 02/12/11
Internet address	http://www.wikicfp.com/cfp/servlet/event.showcfp?eventid=16402&copyownerid=3591

Cite this

@inproceedings{9822196d55fc451d8ec6824d6b1961a8,

title = "Feature selection using misclassification counts",

abstract = "Dimensionality reduction of the problem space through detection and removal of variables, contributing little or not at all to classification, is able to relieve the computational load and instance acquisition effort, considering all the data attributes accessed each time around. The approach to feature selection in this paper is based on the concept of coherent accumulation of data about class centers with respect to coordinates of informative features. Ranking is done on the degree to which different variables exhibit random characteristics. The results are being verified using the Nearest Neighbor classifier. This also helps to address the feature irrelevance and redundancy, what ranking does not immediately decide. Additionally, feature ranking methods from different independent sources are called in for the direct comparison.",

keywords = "Classification, Feature ranking, data reduction, optimisation",

author = "Adil Bagiriv and Andrew Yatsko and Andrew Stranieri and Herbert Jelinek",

year = "2011",

language = "English",

volume = "121",

publisher = "Australian Computer Society Inc",

pages = "51--62",

editor = "Kennedy, {P Kennedy P}",

booktitle = "Proceedings of the Ninth Australasian Data Mining Conference (AusDM 11)",

address = "Australia",

note = "The 9th Australasian Data Mining Conference : AusDM 2011 ; Conference date: 01-12-2011 Through 02-12-2011",

url = "http://www.wikicfp.com/cfp/servlet/event.showcfp?eventid=16402&copyownerid=3591",

}

Feature selection using misclassification counts. / Bagiriv, Adil; Yatsko, Andrew; Stranieri, Andrew et al.
Proceedings of the Ninth Australasian Data Mining Conference (AusDM 11). ed. / P Kennedy P Kennedy. Vol. 121 Sydney, Australia: Australian Computer Society Inc, 2011. p. 51-62.

Research output: Book chapter/Published conference paper › Conference paper › peer-review

TY - GEN

T1 - Feature selection using misclassification counts

AU - Bagiriv, Adil

AU - Yatsko, Andrew

AU - Stranieri, Andrew

AU - Jelinek, Herbert

PY - 2011

Y1 - 2011

N2 - Dimensionality reduction of the problem space through detection and removal of variables, contributing little or not at all to classification, is able to relieve the computational load and instance acquisition effort, considering all the data attributes accessed each time around. The approach to feature selection in this paper is based on the concept of coherent accumulation of data about class centers with respect to coordinates of informative features. Ranking is done on the degree to which different variables exhibit random characteristics. The results are being verified using the Nearest Neighbor classifier. This also helps to address the feature irrelevance and redundancy, what ranking does not immediately decide. Additionally, feature ranking methods from different independent sources are called in for the direct comparison.

AB - Dimensionality reduction of the problem space through detection and removal of variables, contributing little or not at all to classification, is able to relieve the computational load and instance acquisition effort, considering all the data attributes accessed each time around. The approach to feature selection in this paper is based on the concept of coherent accumulation of data about class centers with respect to coordinates of informative features. Ranking is done on the degree to which different variables exhibit random characteristics. The results are being verified using the Nearest Neighbor classifier. This also helps to address the feature irrelevance and redundancy, what ranking does not immediately decide. Additionally, feature ranking methods from different independent sources are called in for the direct comparison.

KW - Classification

KW - Feature ranking, data reduction, optimisation

M3 - Conference paper

VL - 121

SP - 51

EP - 62

BT - Proceedings of the Ninth Australasian Data Mining Conference (AusDM 11)

A2 - Kennedy, P Kennedy P

PB - Australian Computer Society Inc

CY - Sydney, Australia

T2 - The 9th Australasian Data Mining Conference

Y2 - 1 December 2011 through 2 December 2011

ER -

Feature selection using misclassification counts

Abstract

Publication series

Conference

Fingerprint

Cite this