An improved unsupervised clustering algorithm based on population Markov chain

Authors:
F. W. Yang;H. J. Lin;S. H. Yen
Affiliations:
National Tam-Shui Vocational High School;Tamkang University, Taipei, Taiwan, ROC;Tamkang University, Taipei, Taiwan, ROC
Venue:
International Journal of Computers and Applications
Year:
2007

Citing 10
Cited 0

Genetic algorithms + data structures = evolution programs (2nd, extended ed.)

Genetic algorithms + data structures = evolution programs (2nd, extended ed.)
An introduction to genetic algorithms

An introduction to genetic algorithms
Practical genetic algorithms

Practical genetic algorithms
Data clustering: a review

ACM Computing Surveys (CSUR)
A Modified Version of the K-Means Algorithm with a Distance Based on Cluster Symmetry

IEEE Transactions on Pattern Analysis and Machine Intelligence
Global Convergence of Genetic Algorithms: A Markov Chain Analysis

PPSN I Proceedings of the 1st Workshop on Parallel Problem Solving from Nature
Pattern Classification (2nd Edition)

Pattern Classification (2nd Edition)
Pattern Recognition, Third Edition

Pattern Recognition, Third Edition
Some new indexes of cluster validity

IEEE Transactions on Systems, Man, and Cybernetics, Part B: Cybernetics
Degree of population diversity - a perspective on premature convergence in genetic algorithms and its Markov chain analysis

IEEE Transactions on Neural Networks

Quantified Score

Hi-index	0.00

Visualization

Abstract

GA-based clustering approaches have the advantage of automatically determining the optimal number of clusters. In a previous work, we proposed an efficient GA-based clustering algorithm, the PMCC method, and compared it with a representative GA-based clustering algorithm, the GCUK method, to prove its efficiency and effectiveness. In this paper we modify this PMCC method to obtain an improved version: the WPMCC method. This modification prevents premature convergence problem caused in the PMCC method while maintaining the advantage of the PMCC method. The experimental results show that the proposed algorithm not only solves the problem of premature convergence, thereby providing a more stable clustering performance in terms of number of clusters and clustering results, but it also improves the efficiency in terms of time.