Fuzzy models and modeling beyond Artificial Intelligence

In a great variety of application fields where decision making is based on a very large number of factors and/or vague or imprecise conditions it is necessary to introduce some new models and modeling techniques which go beyond the more traditional approaches offered by symbolic Artificial Intelligence. The key problem here is the combination of imprecision and uncertainty with high complexity which cannot be managed by any algorithm with tractable computational complexity not even within the limits of polynomiality [1]. A large portion of decision problems within management engineering falls into this class. In this paper a fuzzy approach is proposed which goes beyond traditional Artificial Intelligence using fuzzy rule based inference systems, while it utilizes other techniques, evolutionary algorithms, neural networks and traditional optimization along with their respective combinations. We hope the proposed tools will be useful for solving some of the real life problems in this area.