M. Jain and M.P. Singh (India)
: Pattern Classification, Artificial NeuralNetwork, Fuzzy Logic
: Artificial neural networks become incompetent when the input information involves fuzziness. It is very difficult for a neural network system to classify a fuzzy input that belongs in two groups or classes with some degree of membership. Hence, the appropriate classification for the fuzzified input pattern can't be competently performed with artificial neural networks. Here, we are providing a resolution for such types of problems by designing a fuzzy neural network controller system. This neuro-fuzzy system consists of two neural network subsystems. First neural network subsystem will be trained for fuzzy control rules. This trained network will produce a consequent for the antecedent of the given fuzzy input control rule and the second neural network subsystem will determine the generalized degree of membership for this consequent. Finally, the neuro-fuzzy controller system will classify it into one of the two classes.
Important Links:
Go Back
