Evaluation of Partitional and Hierarchical Clustering Techniques

Machine learning algorithms were broadly classified into supervised, unsupervised and semi-supervised learning algorithms. Supervised learning algorithms were classified into classification and regression techniques whereas unsupervised learning algorithms were classified into clustering and dimensionality reduction. This paper deals with the evaluation of clustering techniques under unsupervised learning. Clustering is the process of coordinating the data of similar properties under single group. There are several clustering techniques available such as partitional clustering, hierarchical clustering, Fuzzy clustering, Density-based clustering, and Model-based clustering. This paper focuses on the analysis and evaluation of K-means clustering of partitional method and Divisive clustering of hierarchical method. The result of evaluation shows that K-means clustering can hold better for large datasets and it also takes less time than hierarchical clustering.