A prototype of neuro-fuzzy cooperation system

The authors are developing a prototype of a neuro-fuzzy cooperation system that has the precision and learning ability of a neural network and is easy to understand like a fuzzy model. To help convert between neural and fuzzy systems, this system has a neural network with a structure corresponding to that of a fuzzy model. Knowledge acquired from experts was converted from a fuzzy system to a neural network. The neural network was applied to a target system and learned from data obtained during operation to enhance the accuracy of the model. Converting the neural network back into a fuzzy model helps explain the inner representation of the neural network. The model of the target system will be constructed as basic rules and will be improved step by step using a repetition of the fuzzy-neuro and neuro-fuzzy conversion.<<ETX>>