On a new measure of classifier competence applied to the design of multiclassifier systems

This paper presents a new method for calculating competence of a classifier in the feature space. The idea is based on relating the response of the classifier with the response obtained by a random guessing. The measure of competence reflects this relation and rates the classifier with respect to th...

Full description

Bibliographic Details
Main Authors: Woloszynski, Tomasz, Kurzynski, M.
Format: Conference Paper
Published: 2009
Online Access:http://hdl.handle.net/20.500.11937/16628
_version_ 1848749230931312640
author Woloszynski, Tomasz
Kurzynski, M.
author_facet Woloszynski, Tomasz
Kurzynski, M.
author_sort Woloszynski, Tomasz
building Curtin Institutional Repository
collection Online Access
description This paper presents a new method for calculating competence of a classifier in the feature space. The idea is based on relating the response of the classifier with the response obtained by a random guessing. The measure of competence reflects this relation and rates the classifier with respect to the random guessing in a continuous manner. Two multiclassifier systems representing fusion and selection strategies were developed using proposed measure of competence. The performance of multiclassifiers was evaluated using five benchmark databases from the UCI Machine Learning Repository and Ludmila Kuncheva Collection. Classification results obtained for three simple fusion methods and one multiclassifier system with selection strategy were used for a comparison. The experimental results showed that, regardless of the strategy used by the multiclassifier system, the classification accuracy has increased when the measure of competence was employed. The improvement was most significant for simple fusion methods (sum, product and majority vote). For all databases, two developed multiclassifier systems produced the best classification scores. © 2009 Springer Berlin Heidelberg.
first_indexed 2025-11-14T07:17:38Z
format Conference Paper
id curtin-20.500.11937-16628
institution Curtin University Malaysia
institution_category Local University
last_indexed 2025-11-14T07:17:38Z
publishDate 2009
recordtype eprints
repository_type Digital Repository
spelling curtin-20.500.11937-166282017-09-13T13:36:44Z On a new measure of classifier competence applied to the design of multiclassifier systems Woloszynski, Tomasz Kurzynski, M. This paper presents a new method for calculating competence of a classifier in the feature space. The idea is based on relating the response of the classifier with the response obtained by a random guessing. The measure of competence reflects this relation and rates the classifier with respect to the random guessing in a continuous manner. Two multiclassifier systems representing fusion and selection strategies were developed using proposed measure of competence. The performance of multiclassifiers was evaluated using five benchmark databases from the UCI Machine Learning Repository and Ludmila Kuncheva Collection. Classification results obtained for three simple fusion methods and one multiclassifier system with selection strategy were used for a comparison. The experimental results showed that, regardless of the strategy used by the multiclassifier system, the classification accuracy has increased when the measure of competence was employed. The improvement was most significant for simple fusion methods (sum, product and majority vote). For all databases, two developed multiclassifier systems produced the best classification scores. © 2009 Springer Berlin Heidelberg. 2009 Conference Paper http://hdl.handle.net/20.500.11937/16628 10.1007/978-3-642-04146-4_106 restricted
spellingShingle Woloszynski, Tomasz
Kurzynski, M.
On a new measure of classifier competence applied to the design of multiclassifier systems
title On a new measure of classifier competence applied to the design of multiclassifier systems
title_full On a new measure of classifier competence applied to the design of multiclassifier systems
title_fullStr On a new measure of classifier competence applied to the design of multiclassifier systems
title_full_unstemmed On a new measure of classifier competence applied to the design of multiclassifier systems
title_short On a new measure of classifier competence applied to the design of multiclassifier systems
title_sort on a new measure of classifier competence applied to the design of multiclassifier systems
url http://hdl.handle.net/20.500.11937/16628