Clustering based minimum spanning tree algorithm

Abstract

Data mining is a technique used to process information from a big dataset and converting it into a reasonable form for supplementary use. Clustering is a mining technique used in data mining. The goal of clustering is to discover the groupings of a set of points, patterns or objects. Minimum Spanning Tree (MST) based clustering algorithms are successfully used for detecting clusters. In this paper we have focused on minimizing the time complexity for constructing MST by using clustering. The proposed algorithm tries to minimize the time complexity by constructing a MST in two stages. In divide stage, the given dataset is divided in various clusters. In the conquer stage, for every cluster, local MSTs are created and then these MSTs are combined to obtain the final MST by using Midpoint MST algorithm. Experimental results show that the proposed MST algorithm is computationally efficient.