The theory of N-F networks is based on the theory of fuzzy inference (Zadeh, 1973). The difference between fuzzy systems and neuro-fuzzy systems is mainly in the formal representation of the system structure and the methods of tuning the parameters. The neuro-fuzzy system is viewed in the form of a topology similar to the topology of the neural network, there are no equations describing the inference mechanism and there is no list of rules – these elements of the fuzzy system are coded in the form of elements of the network (Rutkowska, 1997; Korbicz and Kowal, 2001). Of course, it is a simple transformation and it does not change the properties of fuzzy inference, it is only another form of realization for the fuzzy system. The crucial difference between fuzzy systems and neuro-fuzzy systems is in the algorithms used to generate the rule base.