Adjusting fuzzy weights in fuzzy neural nets

Abstract

In order to train fuzzy neural nets fuzzy number weights have to be adjusted. Because fuzzy arithmetic automatically leads to monotonic increasing outputs a direct fuzzification of the backpropagation method does not work. Therefore, the focus is on other strategies like evolutionary algorithms. In this paper we suggest a backpropagation based method of adjusting the weights. Furthermore we can show that for the proposed method convergence can be guaranteed.