Discovery by Genetic Algorithm of Cellular Automata Rules for Pattern Reconstruction Task

This paper presents results of the study on application of two-dimensional, three-state cellular automata with von Neumann neighborhood to perform pattern reconstruction task. Searching efficient cellular automata rules is conducted with use of a genetic algorithm. Experiments show a very good performance of discovered rules in solving the reconstruction task despite minimum radius of neighborhood and only partial knowledge about neighborhood states available. The paper also presents interesting reusability possibilities of discovered rules in reconstructing patterns different but similar to ones used during artificial evolution.