Using Multi wavelet Network for Identification

Abstract

Multiwavelet neural network is a new class of artificial neural networks whose their activation functions are multiwavelet functions. The neurons in the hidden layers have been divided into two parts, each part contains wavelets of the same type. The multiwavenet algorithm consist of self-construction of the network and the minimization of the error. Here the constructed network is used for identification of a non-linear function. The network showed the excellent ability to identify the required function compared with the single wavelet network.