Neural Network Design by Using Taguchi Method

Abstract

One of the major difficulties in neural network applications is the selection of the parameters in network configuration and the coefficients in learning rule for fast convergence. This paper develops a network design by combining the Taguchi method and the back-propagation network with an adaptive learning rate for minimum training time and effective vibration suppression. Analyses and experiments show that the optimal design parameters can be determined in a systematic way thereby avoiding the lengthy trial-and-error.