Genetic Design of Independent Input Rule-Based Fuzzy Neural Networks

In this paper, we introduce the genetic design of independent input rule-based fuzzy neural networks. The premise part of the rules of the proposed networks is realized by partitioning of the independent input space using hard-c means clustering. The independently partitioned spaces express the fuzzy rules for respective inputs. The consequence part of the rules is represented by polynomial functions. And the proposed networks are optimized using realcoded genetic algorithms to find the structure and estimate the parameters of the proposed networks. The proposed networks are evaluated the validity using numerical example for nonlinear process.