Model-based architecture for evolutionary intelligent systems

摘要

The basic objective of this research is to develop an architecture for systems capable of evolutionary intelligent behavior by combining the local individual learning within the generation with the global evolution over generations. Thus, the systematic integration of the individual learning with evolutionary learning is our major challenge. To do this, we have proposed the model-based architecture for designing evolutionary intelligent systems on the basis of the advanced modeling and simulation framework. Our approach is compared with conventional intelligent system approaches in that (i) individual and evolutionary intelligences are coherently integrated, and (ii) the modeling and simulation methodology for designing artificial life systems is introduced. The intelligent card game player system has been successfully implemented and tested 10 demonstrate die evolutionary intelligent behavior of proposed architecture.