A Neural Network Approximation of L-MCRS Dynamics for Reinforcement Learning Experiments.
José Manuel López-GuedeManuel GrañaJose Antonio Ramos-HernanzFernando OterinoPublished in: IWINAC (2) (2013)
Keyphrases
- neural network
- reinforcement learning
- recurrent networks
- approximation methods
- function approximators
- function approximation
- artificial neural networks
- dynamic model
- learning capabilities
- pattern recognition
- back propagation
- markov decision processes
- approximation algorithms
- state space
- optimal policy
- dynamical systems
- nonlinear dynamics
- error bounds
- neural network model
- learning algorithm
- bp neural network
- neural network is trained
- genetic algorithm
- multi layer
- training algorithm
- temporal difference
- multilayer perceptron
- feed forward
- knn
- learning process
- multi agent