Dynamic behaviour forecast as a driving force in the coevolution of one-dimensional cellular automata

Various evolutionary methods have been used to look for cellular automata (CA) with a predefined computational behaviour. The most widely studied CA task is the density classification task (DCT) and the best rule currently known for it was obtained by a coevolutionary genetic algorithm (CGA). Here, we analyse the influence of incorporating a parameter-based heuristic into the coevolutionary search. The results obtained show that the parameters can effectively help a CGA in searching for DCT rules, and suggest that the choice of the amount of bias in the search, allowed for the heuristic, is more sensitive than in previous uses we made of it within standard evolutionary algorithms.