Evolving Solutions of the Density Classification Task in 1D Cellular Automata, Guided by Parameters that Estimate their Dynamic Behaviour

Various studies in the Context of one-dimensional cellular automata (CA) have been done on defining parameters directly obtained from their transition rule, which might be able to help forecast their dynamic behaviour. Out of a critique of the most important parameters available for (his end, and out of the definition of a set of guidelines (hat should be followed when defining that kind of parameter, we took two parameters from the literature and defined three new ones, which, jointly provide a good forecasting Set. We then used them to define an evolutionary search heuristic to evolve CA that perform a predefined computational task; here the well-known Density Classification Task is used as reference. The results obtained show that the parameters are effective n helping forecast the dynamic behaviour of onedimensional CA, and can effectively help a genetic algorithm in searching for CA of a predefined kind