A new data clustering using multi-agent turf system

Abstract

In this paper, we proposed a new data clustering method MATS that was inspired by animals, for example, the dog or the ant, which use scent to mark out their turf. MATS can automatically find clusters, depending on a few parameters that are not directly related to the data set. In addition, there are some existence technique was also utilized in our method, such as the density concept and cluster validity index (DB-index). The experiment results verified that MATS is able to discover clusters with varying shapes and is effective when applied to image segmentation.