Effect of the initial configuration of weights on the training and function of artificial neural networks.
R. J. JesusMário Luis Pinto AntunesR. A. da CostaSergey N. DorogovtsevJ. F. F. MendesRui L. AguiarPublished in: CoRR (2020)
Keyphrases
- artificial neural networks
- feedforward artificial neural networks
- multi layer perceptron
- weighted sum
- neural network
- feed forward
- evolutionary artificial neural networks
- output layer
- hidden neurons
- linear combination
- computational intelligence
- feed forward neural networks
- supervised learning
- training phase
- feed forward artificial neural networks
- training speed
- using artificial neural networks
- training process
- multilayer perceptron
- training set
- hidden layer
- neural network model
- activation function
- radial basis function
- genetic algorithm ga
- test set
- weight function
- back propagation
- weight matrix
- training samples
- application of artificial neural networks
- least squares
- evolutionary algorithm