MULTIPLE INTERACTING PROGRAMS - A REPRESENTATION FOR EVOLVING COMPLEXBEHAVIOR

This paper defines a representation for expressing complex behaviors c alled multiple interacting programs (MIPs) and describes an evolutiona ry method for evolving solutions to difficult problems expressed as MI Ps structures. The MIPs representation is a generalization of neural n etwork architectures that can model any type of dynamic system. The ev olutionary training method described is based on an evolutionary progr am originally used to evolve the architecture and weights of recurrent neural networks. Example experiments demonstrate the training method' s ability to evolve appropriate MIPs solutions for difficult problems. An analysis of the evolved solutions shows their dynamics to be inter esting and nontrivial.