Evolving the Topology and the Weights of Neural Networks Using a Dual Representation
Applied Intelligence
Training a neuron in sequential learning
Neural Processing Letters
A Comparison of Methods for Learning of Highly Non-separable Problems
ICAISC '08 Proceedings of the 9th international conference on Artificial Intelligence and Soft Computing
Hi-index | 0.01 |
We propose an efficient procedure for constructing and traininga feedforward neural network. The network can perform binaryclassification for binary or analogue input data. We show that theprocedure can also be used to construct feedforward neural networkswith binary-valued weights. Neural networks with binary-valuedweights are potentially straightforward to implement usingmicroelectronic or optical devices and they can also exhibit goodgeneralization.