원문정보
초록
영어
This paper introduce neuro-fuzzy networks by means of fuzzy relation space-based Rules for pattern recognizer. The proposed neuro-fuzzy networks are realized with the aid of the grid partition of the fuzzy relation input space. The partitioned spaces express the fuzzy rules of the networks. The consequence part of the rules is represented by polynomial functions whose coefficients are learned by the back-propagation algorithm. To optimize the parameters of the proposed networks, we consider real-coded genetic algorithms. The proposed networks are evaluated with the use of numerical experimentation for pattern recognizer. Finally, this paper shows that the proposed networks have the good result.
목차
1. Introduction
2. Design of the fuzzy relation-based neuro-fuzzy networks
2.1. The structure of the fuzzy relation-based NFNs
2.2 The learning algorithm
3. Genetic optimization
4. Experimental Studies
5. Conclusions
References