Using a Hybrid Cellular Automata Topology and Neighborhood in Rule Discovery

Cellular Automata are important tools in the study of complex interactions and analysis of emergent behaviour. They have the ability to generate highly complex behaviour starting from a simple initial configuration and set of update rules. Finding rules that exhibit a high degree of self-organization is a challenging task of major importance in the study of complex systems. In this paper, we propose a new cellular automaton (CA) topology and neighbourhood that can be used in the discovery of rules that trigger coordinated global information processing. In the introduced approach, the state of a cell changes according to the cell itself, the cells in the local neighborhood as well as some fixed long-distance cells. The proposed topology is engaged to detect new rules using an evolutionary search algorithm for the well-known density classification task. Experiments are performed for the one-dimensional binary-state CA and results indicate a good performance of the rules evolved by the proposed approach.