A novel clustering algorithm based on P systems

Abstract

Membrane computing (known as P systems) is a novel class of distributed parallel computing models. In this paper, a partition-based clustering algorithm under the framework of membrane computing is proposed. The clustering algorithm is based on a tissue-like P system, which is used to exploit the optimal cluster centers for a data set. Each object in the tissue-like P system represents a group of candidate cluster centers and is evolved through simulated annealing mechanism and mutation mechanism. Meanwhile, communication rules are used to exchange and share the objects between different elementary membranes and between elementary membranes and the environment. The proposed clustering algorithm is evaluated over two artificial data sets and two real-life data sets and is further compared with k-means algorithm and GA-based k-means algorithm respectively. The comparison results reveal the superiority of the proposed clustering algorithm in terms of clustering quality and stability