Abstract:
Clustering is a widely used unsupervised data analysis technique in machine learning. However, a common requirement amongst many existing clustering methods is that all pairwise distances between patterns must be computed in advance. This makes it computationally expensive and difficult to cope with large scale data used in several applications, such as in bioinformatics. In this paper we propose a novel sequential hierarchical clustering technique that initially builds a hierarchical tree from a small fraction of the entire data, while the remaining data is processed sequentially and the tree adapted constructively. Preliminary results using this approach show that the quality of the clusters obtained does not degrade while reducing the computational needs.
