Fault tolerant training algorithm for multi-layer neural networks focused on hidden unit activities.
Haruhiko TakaseHidehiko KitaTerumine HayashiPublished in: IJCNN (2006)
Keyphrases
- fault tolerant
- multi layer
- training algorithm
- neural network
- back propagation
- fault tolerance
- neural nets
- feed forward neural networks
- neural network training
- distributed systems
- error back propagation
- artificial neural networks
- fuzzy logic
- load balancing
- training process
- multilayer perceptron
- pattern recognition
- feed forward
- single layer
- bp neural network
- hidden layer
- recurrent neural networks
- neural network model
- activation function
- prediction model
- feature vectors
- decision making
- genetic algorithm