A dynamic genetic clustering algorithm for automatic choice of the number of clusters

One of the most difficult problems in cluster analysis is how many clusters are appropriate for the description of a given system. In this paper, a novel dynamic genetic clustering algorithm (DGCA) is proposed to automatically search for the best number of clusters and the corresponding partitions. In the DGCA, a maximum attribute range partition approach is used in the population initialization in order to overcome the sensitivity of clustering algorithms to initial partitions. Furthermore, the methods of two-step selection and mutation operations are developed to exploit the search capability of the algorithm. Finally, the comparison among the DGCA, k-means algorithm and the standard genetic k-means clustering algorithm (SGKC) is illustrated with several artificial and real life data sets.