A New Algorithm for Structure Optimization of Wavelet Neural Network
This paper presents a new algorithm for constructing and training wavelet neural network. This algorithm is based on the variation of the number of hidden neurons dynamically during the training process. The suggested method determines the optimal number of the hidden neurons and solves the optimization problem of wavelet neural network structure. The problem of finding a good neural model is then discussed through solutions offered by wavelet neural networks trained by conjugate gradient algorithm. Finally, experimental results, which confirm the efficiency of this approach, are reported.
Keywords: Neural Network, Wavelet, Conjugate Gradient, Regressor, Local Minimum, Hidden Neuron
Download Full-Text
