Robust Reinforcement Learning Control Using Integral Quadratic Constraints for Recurrent Neural Networks.
Charles W. AndersonPeter Michael YoungMichael R. BuehnerJames N. KnightK. A. BushDouglas C. HittlePublished in: IEEE Trans. Neural Networks (2007)
Keyphrases
- recurrent neural networks
- reinforcement learning
- feed forward
- control problems
- optimal control
- neural network
- control system
- robot control
- feedforward neural networks
- reservoir computing
- artificial neural networks
- echo state networks
- recurrent networks
- action selection
- complex valued
- control policy
- objective function
- function approximation
- neural model
- nonlinear dynamic systems
- model free
- adaptive control
- optimal policy
- state space
- learning algorithm