Synchronization learning of coupled chaotic maps

We study the dynamics of an ensemble of globally coupled chaotic logistic m aps under the action of a learning algorithm aimed at driving the system fr om incoherent collective evolution to a state of spontaneous full synchroni zation. Numerical calculations reveal a sharp transition between regimes of unsuccessful and successful learning as the algorithm stiffness grows. In the regime of successful learning, an optimal value of the stiffness is fou nd for which the learning time is minimal.