Cellular Automata based Adaptive Clustering Approach

Clustering is a popular machine learning approach, and research on clustering is always a hot spot. A cellular automata based clustering is proposed in this paper. Cellular automata is introduced to model the comparison between entities and the forming process of the clustered groups. In this approach, entities that needs clustering could be placed into the different cells in the initial phase. In the clustering forming process, the entities in a specific cell could compare with the entities in the neighboring cells. The next action of entities in the cell could be made according to the situation in and around the cell. The similar entities would be moved to form groups. Simulation and experiments demonstrate the effectiveness of the proposed approach.