The roles of soft computing and fuzzy logic in the conception, design and deployment of intelligent systems

Summary form only given. Soft computing (SC) is a consortium of methodologies which provide a foundation for intelligent systems. The principal methods are fuzzy logic (FL), neurocomputing (NC), genetic computing (GC) and probabilistic computing (PC), with PC subsuming evidential reasoning, uncertainty management and some machine learning theory. The main contribution of FL is a methodology for dealing with imprecision, approximate reasoning, fuzzy information granulation and computing with words; that of NC system identification, learning and adaption; that of CC systematized random research, tuning and optimization; and that of PC decision analysis and uncertainty management. The guiding principle of soft computing is: Exploit the tolerance for imprecision, uncertainty and partial truth to achieve tractability, robustness, low solution cost and better rapport with reality. The 4 methods are complementary rather than competitive. Their use in combination leads to hybrid intelligent systems. The most visible of such systems are neuro-fuzzy systems. The ubiquity of intelligent systems is certain to have a profound impact on the ways in which man-made systems are conceived, designed, manufactured, employed and interacted with. This is the perspective in which basic issues relating to soft computing and intelligent systems are addressed.