A-2-15 Learning Ability of Evolving Affordable Neural Networks with Genetic Algorithm for Back Propagation Learning

In our previous research, we have proposed a new network structure with affordable neurons in hidden layer of the feedforward neural network. We named this proposed network “Affordable Neural Network” [1]. We consider that the operation of the affordable neural network is especially effective on evolution of neural network in brain. In this study, we investigate the performance of the affordable neural network when the hidden layer is evolved using GA. By computer simulations, we consider that the affordable neurons exert an important influence on evolution process in the hidden layer of the network.

[1]  Yoshifumi Nishio,et al.  Back propagation learning of neural networks with chaotically-selected affordable neurons , 2005, 2005 IEEE International Symposium on Circuits and Systems.