Combining multiple neural network paradigms and applications using SESAME

SESAME (software environment for the simulation of adaptive modular system) has been developed to make experiments possible in fields like pattern recognition and control that combine different neural network models and learning paradigms in an elegant manner. SESAME represents an object-oriented integrated framework for experiments in pattern recognition and control. All experiment components are uniform building blocks, which can be put together arbitrarily by means of a common communication interface. Experiment substructures can be integrated to construct building blocks, which may be reused for other experiments. Examples drawn from different application domains and learning paradigms illustrate the potential of this approach