Incremental Clustering for Hierarchical Clustering

Abstract

This paper proposes a clustering algorithm for updating clusters without reclustering when a point is inserted. We define the center and the radius of the cluster, and update clustering results of points using them. We introduce the concept of outliers and also consider the change in the number of clusters caused by data insertion. From comparative experiments with reclustering by the conventional method, it is shown that the proposed method can cluster points with short calculation time.