Identification of Cellular Automata by Immunocomputing
Alexander Tarakanov and Alexander Prokaev
This paper proposes a new method of identification of cellular automata, using immunocomputing. The essence of the approach is the representation of states and transitions of the automaton by using a formal immune network and the faultless reducing of number of the transitions by apoptosis and immunization. Numerical examples utilize excitable media and real-world atlas of the sea surface temperature and demonstrate the advantage of our approach compared to neurocomputing and conventional interpolation.