Genetic algorithm-based learning of fuzzy neural networks. Part 1: feed-forward fuzzy neural networks

In spite of great importance of fuzzy feed-forward and recurrent neural networks (FNN) for solving wide range of real-world problems, today there is no effective learning algorithm for FNN. In this paper we propose an effective genetic-based learning mechanism for FNN with fuzzy inputs, fuzzy weights expressed as LR-fuzzy numbers, and fuzzy outputs. The effectiveness of the proposed method is illustrated through simulation of fuzzy regression for quality evaluation and comparison with the widely used learning method based on a-cuts and fuzzy arithmetic. Finally, we demonstrate the use of the proposed learning procedure for calculating fuzzy-valued profit in an oligopolistic environment. (C) 2001 Elsevier Science B.V, All rights reserved.