Weight Initialization for Simultaneous Recurrent Neural Network Trained with a Fixed point Learning Algorithm.
Gürsel SerpenYifeng XuPublished in: Neural Process. Lett. (2003)
Keyphrases
- fixed point
- recurrent neural networks
- learning algorithm
- echo state networks
- initial conditions
- neural network structure
- feed forward
- neural network
- back propagation
- training process
- recurrent networks
- hidden layer
- complex valued
- sufficient conditions
- multilayer perceptron
- artificial neural networks
- dynamical systems
- machine learning
- belief propagation
- reservoir computing
- training data
- multi layer perceptron
- floating point
- graphical models
- reinforcement learning
- constraint databases
- least squares
- rbf network
- policy iteration
- learning rate
- training algorithm
- knowledge base
- decision making
- genetic algorithm
- learning tasks