Welcome to the InfoSci Platform
Reference Hub3
GeneticTKM: A Hybrid Clustering Method Based on Genetic Algorithm, Tabu Search and K-Means

GeneticTKM: A Hybrid Clustering Method Based on Genetic Algorithm, Tabu Search and K-Means

Masoud Yaghini, Nasim Gereilinia
Copyright: © 2013 |Volume: 4 |Issue: 1 |Pages: 11
ISSN: 1947-8283|EISSN: 1947-8291|EISBN13: 9781466631182|DOI: 10.4018/jamc.2013010105
Cite Article Cite Article

MLA

Yaghini, Masoud, and Nasim Gereilinia. "GeneticTKM: A Hybrid Clustering Method Based on Genetic Algorithm, Tabu Search and K-Means." IJAMC vol.4, no.1 2013: pp.67-77. http://doi.org/10.4018/jamc.2013010105

APA

Yaghini, M. & Gereilinia, N. (2013). GeneticTKM: A Hybrid Clustering Method Based on Genetic Algorithm, Tabu Search and K-Means. International Journal of Applied Metaheuristic Computing (IJAMC), 4(1), 67-77. http://doi.org/10.4018/jamc.2013010105

Chicago

Yaghini, Masoud, and Nasim Gereilinia. "GeneticTKM: A Hybrid Clustering Method Based on Genetic Algorithm, Tabu Search and K-Means," International Journal of Applied Metaheuristic Computing (IJAMC) 4, no.1: 67-77. http://doi.org/10.4018/jamc.2013010105

Export Reference

Mendeley
Favorite Full-Issue Download

Abstract

The clustering problem under the criterion of minimum sum square of errors is a non-convex and non-linear problem, which possesses many locally optimal values, resulting that its solution often being stuck at locally optimal solution. In this paper, a hybrid genetic, tabu search and k-means algorithm, called GeneticTKM, is proposed for the clustering problem. A new mutation operator is presented based on tabu search algorithm for the proposed hybrid genetic method. The key idea of the new operator is to produce tabu space for escaping from trap of local optimal and finding better solution. The results of the proposed algorithm are compared with other clustering algorithms such as genetic algorithm; tabu search and particle swarm optimization by implementing them and using standard and simulated data sets. The authors also compare the results of the proposed algorithm with other researchers’ results in clustering the standard data sets. The results show that the proposed algorithm can be considered as an effective and efficient algorithm to find better solution for the clustering problem.

Request Access

You do not own this content. Please login to recommend this title to your institution's librarian or purchase it from the IGI Global bookstore.