Towards formal theory of soft computing

Abstract

 Soft computing (SC) is usually presented as a new computational paradigm which occurs when mixing the methods of fuzzy logic, neural networks and genetic algorithms to obtain more sophisticated results while solving problems under imprecision and uncertainty. In my opinion, SC is now in a stage when it is possible to establish it as a theory with clearly specified subject and to specify its language and methods. Few papers discussing various aspects of it have been published (Bonissone 1997, Pap 1997, Zadeh 1997, Wang and Shaohua 1997). All of them demonstrated that SC is especially a theory open to various sources, not limited only to the above three mentioned ones.