ANALYSIS OF HIERARCHICAL CLUSTERING ALGORITHM TO HANDLE LARGE DATASET

Abstract

Clustering, in Data Mining is useful for discovering groups and identifying interesting distributions in
underlying data. Traditional data clustering algorithms either favor clusters with special shapes and similar sizes, or are
very delicate in the presence of outliers. Nowadays most widely studied problem is identification of clusters in a large
dataset. Hierarchical Clustering is the process of forming a maximal collection of subsets of objects (called clusters),
with the property that any two clusters are either disjoint or nested. Hierarchical clustering combine data objects into
clusters, those clusters into larger clusters, and so forth, creates a hierarchy of clusters, which may represent a tree
structure called a dendrogram. Agglomerative clustering is a most flexible method and it is also used for clustering the
large dataset, there is no need of the number of clusters as an input. In this paper we have introduced solution for
decreasing time complexity of clustering algorithms by combining approaches of two different algorithms from which
one is good in accuracy and other is fast that is helpful for information retrieval from large data.