Acquisition of fuzzy rules using fuzzy neural networks with forgetting
Motohide Umano, Shiro Fukunaka, Itsuo Hatono, Hiroyuki Tamura · Proceedings of International Conference on Neural Networks (ICNN'97) · 2002
We acquire fuzzy rules from data using a fuzzy neural network. First, we generate an initial fuzzy neural network of the specified number of fuzzy rules that have fewer good membership functions than generated using a self-organization algorithm by Kohonen. Then, we tune and prune fuzzy rules based on a structural learning algorithm with forgetting by Ishikawa (1996), where the numerals in the consequent part and the center values and widths of membership functions in the antecedent part are tuned and forgotten a little, and thus redundant rules and variables are pruned to acquire simpler, general rules. We apply the method to the iris classification problem, of Fisher (1936) and have a very good result.