Estimation of size of hidden layer on basis of bound of generalization error.
Boris IgelnikYoh-Han PaoPublished in: ICNN (1995)
Keyphrases
- training error
- hidden layer
- generalization error
- upper bound
- back propagation
- neural network
- feed forward
- learning rate
- learning algorithm
- multilayer perceptron
- recurrent neural networks
- artificial neural networks
- radial basis function
- active learning
- rbf neural network
- classification error
- cross validation
- generalization error bounds
- feedforward neural networks
- activation function
- latent variables
- model selection
- sample size
- error rate
- training algorithm
- lower bound
- estimation error
- neural nets
- training set
- gradient method
- training data
- machine learning
- control system
- probability distribution
- support vector
- bayesian networks
- multi class
- linear combination
- machine learning algorithms
- adaboost algorithm
- feature selection
- minimum margin