Abdullah, , and Ku-Mahamud, Ku Ruhana (2015) Ant system-based feature set partitioning algorithm for K-NN and LDA ensembles construction. In: 5th International Conference on Computing and Informatics (ICOCI) 2015, 11-13 August 2015, Istanbul, Turkey.

Preview

PDF Download (383kB) | Preview

Abstract

Combination of several classifiers has been very useful in improving the prediction accuracy and in most situations multiple classifiers perform better than single classifier.However not all combining approaches are successful at producing multiple classifiers with good classification accuracy because there is no standard resolution in constructing diverse and accurate classifier ensemble.This paper proposes ant system-based feature set partitioning algorithm in constructing k-nearest neighbor (k-NN) and linear discriminant analysis (LDA) ensembles. Experiments were performed on several University California, Irvine datasets to test the performance of the proposed algorithm.Experimental results showed that the proposed algorithm has successfully constructed better classifier ensemble for k-NN and LDA.

Item Type:	Conference or Workshop Item (Paper)
Additional Information:	ISBN No: 978-967-0910-02-4 Joinly organized by: Universiti Utara Malaysia & Istanbul Zaim University
Uncontrolled Keywords:	k-nearest neighbor, linear discriminant analysis, feature set partitioning, ant system algorithm, classifier ensemble
Subjects:	Q Science > QA Mathematics > QA75 Electronic computers. Computer science
Divisions:	School of Computing
Depositing User:	Prof. Dr. Ku Ruhana Ku Mahamud
Date Deposited:	01 Oct 2015 06:31
Last Modified:	26 Apr 2016 08:14
URI:	https://repo.uum.edu.my/id/eprint/15575

Actions (login required)

View Item