IK-BKM: An incremental clustering approach based on intra-cluster distance

This paper introduces a novel incremental approach to clustering uncertain categorical data. This so-called Incremental K Belief K-modes Method (IK-BKM) extends the Belief K-modes one to update the cluster partition when new information is available namely the increase of final desired clusters´ number. The main objective is to update clusters´ partition without complete reclustring. Our method will be illustrated by an example showing the comparative results of the incremental process and the non incremental one.