An approach to pattern recognition by fuzzy category and neural network simulation
Wang-Kun Chen · 2010
This paper presents a new approach to extract the interpretable knowledge from the experimental data. A pattern rule is first generated followed by Bayes' theorem. The pattern was designed by the Bayes' classifier for data clustering. The data from the optimized category of fuzzy system was then transferred to the neural network for refining the obtained knowledge. The optimized fuzzy system could extract the understandable knowledge from the measured results. Different neural network method could be used in the algorithm. Simulation results on the phenomenon show that the approach to explain the natural environment is effective.