Optimal approximation of square integrable functions by a flexible one-hidden-layer neural network of excitatory and inhibitory neuron pairs.
Bertrand GiraudLon Chang LiuChristophe BernardHerbert AxelradPublished in: Neural Networks (1991)
Keyphrases
- hidden layer
- neural network
- activation function
- back propagation
- output layer
- feed forward
- feed forward neural networks
- artificial neural networks
- multilayer perceptron
- recurrent neural networks
- radial basis function
- feedforward neural networks
- single hidden layer
- rbf neural network
- learning rate
- neural nets
- training algorithm
- hidden neurons
- basis functions
- connection weights
- radial basis function neural network
- neural network structure
- number of hidden layers
- rbfnn
- bp neural network
- latent variables
- neural network model
- hidden nodes
- sigmoid function
- backpropagation algorithm
- dynamic programming
- fuzzy neural network
- multi layer
- pattern recognition
- neural model
- knn
- probabilistic model
- genetic algorithm
- learning algorithm
- training set
- single layer
- support vector machine svm
- random variables
- fault diagnosis
- nonlinear functions
- learning rules
- data fusion