A multilayer feedforward fuzzy neural network


Savran A.

ARTIFICIAL INTELLIGENCE AND NEURAL NETWORKS, vol.3949, pp.78-83, 2006 (Journal Indexed in SCI) identifier

  • Publication Type: Article / Article
  • Volume: 3949
  • Publication Date: 2006
  • Title of Journal : ARTIFICIAL INTELLIGENCE AND NEURAL NETWORKS
  • Page Numbers: pp.78-83

Abstract

This paper describes the architecture and learning procedure of a multilayer feedforward fuzzy neural network (FNN). The FNN is designed by replacing the sigmoid type activation function of the multilayer neural network (NN) with the fuzzy system (FS). The Levenberg-Marquardt (LM) optimization method with a trust region approach is adapted to train the FNN. Simulation results of a non-linear system identification problem are given to show the validity of the approach.