The present invention relates to a fuzzy model using a neural network suitable for improving the performance of an inaccurate fuzzy model. In the past, it has been difficult to estimate the coefficients of the fuzzy set and the output function constituting the fuzzy model. A fuzzy model can be constructed by using a neural network circuit capable of repeated learning, and a more accurate fuzzy model can be easily formed, so that the above determination can be improved.
展开▼