A novel approach for hierarchical clustering in non - binary search space

Data clustering is one of the powerful techniques for the knowledge discovery from data. In this paper, a novel approach for hierarchical clustering has been proposed over non-binary search space. Besides the agglomerative methods, the proposed algorithm has considered the Strength of Presence associated with each transaction, to yield quality clusters which are again more close to the real life situation. The algorithm is also capable to produce clusters with customizable varying density along with varying number of clusters.