The project is concerned with developing problem-solving software that is based on genetic algorithms (GA) and classifier- systems (CS). This software models adaptive systems that operate in complex non-linear environments in the following manner: (1) Goal-directed action involves the construction, testing, and use of internal, forward-predicting models of the environment, and the constant adaptation of such models to the changes of the environment. (2) A tag on a classifier give it a recognizable identity, so classifiers can learn to cooperate, as in a forward- predicting model. //
