Bidirectional Neural Networks Reduce Generalization Error.
Akbar Farhoodi NejadTamás D. GedeonPublished in: IWANN (1995)
Keyphrases
- generalization error
- neural network
- cross validation
- model selection
- upper bound
- training error
- active learning
- learning algorithm
- classification error
- training data
- binary classification
- linear classifiers
- training set
- target function
- sample complexity
- sample size
- artificial neural networks
- learning machines
- fuzzy logic
- generalization error bounds
- supervised learning
- multilayer perceptron
- boosting algorithms
- conditional expectation
- training process
- subspace information criterion
- perceptron learning