RLDDE: A novel reinforcement learning-based dimension and delay estimator for neural networks in time series prediction.
F. LiuGeok See NgChai QuekPublished in: Neurocomputing (2007)
Keyphrases
- reinforcement learning
- neural network
- function approximation
- neural network model
- global exponential stability
- function approximators
- artificial neural networks
- recurrent neural networks
- least squares
- mackey glass
- non stationary
- temporal difference
- pattern recognition
- neuro fuzzy
- maximum likelihood
- back propagation
- reinforcement learning algorithms
- learning capabilities
- genetic algorithm
- radial basis function
- markov decision processes
- feed forward
- neural nets
- learning algorithm
- multilayer perceptron
- state space
- recursive least squares
- fuzzy logic
- fault diagnosis
- transfer learning
- fuzzy systems
- model free
- monte carlo
- multi layer perceptron
- action space
- learning classifier systems
- estimation algorithm
- variance estimator
- multi layer