Genetic Algorithms as Heuristics for Optimizing ANN Design

Abstract

The problem of the ANN design is usually thought as residing in solving the training problem for some predefined ANN structure and connectivity. Training methods are very problem and ANN dependent. They are sometimes very accurate procedures but they work in narrow and restrictive domains. Thus the designer is faced to a wide diversity of different training mechanisms. We have selected Genetic Algorithms because of their robustness and their potential extension to train any ANN type. Furthermore we have addressed the connectivity and structure definition problems in order to accomplish a full genetic ANN design. These three levels of design can work in parallel, thus achieving multilevel relationships to build better ANNs. GRIAL is the tool used to test several new and known genetic techniques and operators. PARLOG is the Concurrent Logic Language used for the implementation in order to introduce new models for the genetic work and to attain an intralevel distributed search as well as to parallelize any ANN management and any genetic operations.