A New Categorical Data Clustering Technique Based on Genetic Algorithm

摘要

This paper presents a categorical clustering algorithm based on genetic algorithm. Clusters are represented using summary vectors. Genetic algorithm is used to search for appropriate cluster summary vectors. Chromosomes are of variable lengths and encode the cluster summaries. This algorithm is applied on a number of real life data sets. Experimental results show that this method gives pure clusters on most of the datasets with no errors present. For the other data sets also very small amount of errors have been noticed.