A Comparative Study on K-Means Clustering and Agglomerative Hierarchical Clustering

Clustering is a well-established unsupervised data mining approach that group data points based on similarities. Clustering entities will give insights into the characteristics of different groups. Clustering results in minimization of the dimensionality of data set when you are dealing with a myriad number of data. The higher the homogeneity within the cluster and the higher the differences between the clusters, the finer the cluster will be. Clusters are mainly of two types: 1) Soft clustering: Based on the probability that a data point will belong to a specific cluster and, 2) Hard clustering: Data points are separated into independent clusters. Among hundreds of clustering algorithms, they can be labeled into one of following models such as connectivity, density, distribution and centroid model. This paper attempts to differentiate two widely used clustering techniques, k-means clustering and hierarchical clustering which belong to the centroid and connectivity models respectively. The comparison will be based on execution time and memory usage of both these algorithms when different sets of a delivery fleet driver data set are manipulated using these algorithms.