S. Jafari and S.D. Katebi (Iran)
Fuzzy Expert System, Neural Network, Hybrid System, Back-Propagation, Certainty Factor CF, Neuro-Fuzzy network.
Presented is a description of an algorithm that has been
used for Fuzzy-Certainty Factor expert systems tuning.
This process has been implemented on a Dental Diseases
Diagnostic System (DEXSYS) developed by the authors.
It is attempted to translate or map a fuzzy logic system to
neural network framework and use the neural networks
learning methods for tuning and adaptation of Fuzzy
Certainty Factor expert system based on the back
propagation algorithm. Since the tuning process is an off
line one, the inherent slowness of the back-propagation is
ignorable regarding to other methods such as Kohonen.