Comparison of the K-Means and the K-Medoids Partitioning Algorithms for Clustering of Optical Mineralogy

Abstract

Clustering is the process of grouping the data into classes or clusters so that objects within a cluster have high similarity in comparison to one another, but are very dissimilar to objects in other clusters. The dissimilarity (or similarity) between the objects described by interval-scaled variables is typically computed based on the distance between each pair of objects. Euclidean distance is used in this system. K-Means and k-Medoids algorithms are by far the most widely used method for discovering clusters in data. This system can implement the clustering of optical mineralogy by using the two partitioning methods (k-Means algorithm and k-Medoids algorithm) and then evaluate the performance of the processing time, squared-error rate and average squared-error value of these algorithms. Experiments show that these two algorithms are effective for 200 records with fourteen attributes of optical mineralogy datasets and becomes more and more effective as the number of clusters increases.