A genetic algorithm based clustering approach with Tabu operation and K-means operation

作者：Liu, Y^*; Yan, Hua; Chen, Kefei

来源：Journal of Intelligent Systems, 2010, 19(1): 17-46.

DOI：10.1515/JISYS.2010.19.1.17

• Clustering
• Clustering analysis
• Clustering approach
• Clustering problems
• k-Means algorithm
• Population diversity
• Real life datasets
• Speed of convergence

Clustering analysis is a nonconvex problem that possesses many locally optimal values, with the result that its solution often falls into these traps. In this article, a hybrid genetic clustering algorithm called Genetic algorithm with Tabu operation and AT-means operation based Clustering (GTK-Clustering) is developed to deal with the clustering problem. With the cooperation of tabu operation and k-means operation, the GTKClustering algorithm can keep a balance between population diversity and the speed of convergence. On the one hand, the tabu operation prevents the population from being dominated by several fitter individuals and maintains a high level of population diversity, and on the other hand, the k-means operation improves the distribution of objects and enhances the speed of convergence of the clustering algorithm. Its superiority over the k-means algorithm and another genetic clustering approach is demonstrated for artificial and real life data sets.

• 出版日期2010
• 单位上海交通大学; 电子科技大学