Accelerated gradient learning algorithm for neural network weights update.
Zeljko HocenskiMladen AntunovicDamir FilkoPublished in: Neural Comput. Appl. (2010)
Keyphrases
- neural network
- learning algorithm
- back propagation
- single hidden layer
- connection weights
- hidden neurons
- training algorithm
- generalization ability
- hidden layer
- weight update
- gradient vector
- artificial neural networks
- learning rules
- bp neural network
- neural network structure
- training data
- weighted sum
- feed forward
- neural network model
- rbf network
- learning scheme
- activation function
- feed forward neural networks
- weighted sums
- edge detector
- learning rate
- linear combination
- feedforward neural networks
- neural nets
- extreme learning machine
- fuzzy logic
- pattern recognition
- genetic algorithm
- multilayer perceptron
- radial basis function
- learning problems
- machine learning
- edge detection
- relative importance
- fuzzy artmap
- neural network is trained
- gradient method
- weight vector
- active learning
- machine learning algorithms
- training examples
- supervised learning
- hopfield neural network
- gradient information
- bp algorithm
- hidden nodes
- gradient direction
- nearest neighbour
- weighting scheme
- data sets
- fault diagnosis
- training samples