Performance evaluation results of evolutionary clustering algorithm star for clustering heterogeneous datasets

This article presents the data used to evaluate performance of evolutionary clustering algorithm star (ECA*) compared five traditional and modern algorithms. Two experimental methods are employed examine ECA* against genetic for clustering++ (GENCLUST++), learning vector quantisation (LVQ), expectation maximisation (EM), K-means++ (KM++) K-means (KM). These algorithms applied 32 heterogenous multi-featured datasets determine which one performs well on three tests. For one, ther paper examines efficiency in contradiction its corresponding using evaluation measures. validation criteria objective function cluster quality another, it suggests a rating framework measurethe sensitivity these varos dataset features (cluster dimensionality, number clusters, overlap, shape structure). The contributions experiments two-folds: (i) exceeds counterpart aloriths ability find out right number; (ii) is less sensitive towards competitive techniques. Nonetheless, results performed demonstrate some limitations ECA*: not fully based premise that no prior knowledge exists; Adapting utilising several real applications has been achieved yet.