A hybrid method combining P systems and gravitational search for data clustering

摘要

Membrane computing is a class of distributed parallel computing models. This paper proposes a hybrid approach to process the data clustering problem. A cell P system is used as its computing framework and gravitational search mechanism is combined as evolution rules of objects in cells. Due to parallel computing ability and evolution-communication mechanism of the cell P system, the presented hybrid method can effectively and efficiently find the optimal cluster centers for a data set to be clustered. The performance of the presented hybrid method is studied with several evolutionary algorithms. Simulation results show that the presented hybrid approach is superior or comparable to k-means algorithm and other evolutionary algorithms for data clustering.