Abstracts > Perrotin Pacôme

Automata networks are dynamical systems, and computing the limit cycles and fixed points (the attractors) of these networks directly is exponential in their size. In this talk we present a modular variation of these networks with added inputs. By focusing the acyclic case, we are able to characterise the attractors of a network by the use of output functions. These output functions allow for the description of equivalence classes, but also an optimisation method for the size of networks, a direct and faster way of computing attractors and a transformation of the problem of computing attractors into various other combinatorial problems.