Preface: 3rd international conference on fuzzy logic, neural nets, and soft computing

This is a special issue on fuzzy neural systems. The papers were mainly selected from the presentations in the 3rd International Conference on Fuzzy Logic, Neural Nets, and Soft Computing ( I I Z U K A '94) that was held in Iizuka, August 1-7, 1994. Integration of fuzzy systems and neural systems is an active research area. Many papers were presented in this area at I I Z U K A '94, and six papers were selected after the usual review process. Fuzzy systems and neural systems each have their own merits: Fuzzy systems can utilize linguistic knowledge of human experts in the form of fuzzy if-then rules, and neural systems can be trained by numerical input-output data. Many approaches have been proposed for integrating these two systems. Those approaches can be classified into several categories. One category consists of multilayer feedforward neural networks for fuzzy reasoning. Those neural networks are sometimes referred to as fuzzy neural networks. The first paper, by Horikawa, Furuhashi, and Uchikawa, can be classified in this category. The authors propose a fuzzy neural network architecture based on the truth space fuzzy reasoning approach for automatically acquiring fuzzy rules with linguistic hedge. This architecture is their fifth fuzzy neural network. The second paper, by Lee, Lee, and Park, can be also classified in the same category as the first paper. In it, a new controller design algorithm is proposed based on a neurofuzzy identi-