A Genetic Clustering Method Based on Variable Length String

Abstract

This paper proposes a method based on variable length string genetic algorithm to solve the clustering problem. This method generates the optimal solution by genetic operation of variable length chromosomes, automatically determines the most appropriate number of cluster centers, and effectively solves the existing K-means clustering algorithms’ defects such as highly subject to the constraint of the initial clustering centers and need to specify the specific number of clustering centers. The experimental results and comparison show that the purposed method’s effectiveness.