ニューロ・ファジイ学習アルゴリズムにおける学習係数について(一般講演,医療の将来とソフトコンピューティング)

抄録

In this paper, we study the problem of designing of learning rates in a class of neuro-fuzzy learning algorithms, which are widely used in recent fuzzy applications for tuning fuzzy inference rules. The analysis on the convergence of the learning algorithm in the cases of tuning parameters of fuzzy rules is conducted. Sufficient conditions for choosing appropriate learning rates have been presented such that an optimal fuzzy system model can be constructed.