Model-free approximate dynamic programming for continuous-time linear systems.
Jae Young LeeJin Bae ParkYoon Ho ChoiPublished in: CDC (2009)
Keyphrases
- approximate dynamic programming
- linear systems
- model free
- reinforcement learning
- dynamical systems
- policy iteration
- state space
- optimal control
- function approximation
- reinforcement learning algorithms
- linear program
- temporal difference
- sufficient conditions
- markov chain
- fixed point
- control policy
- markov decision processes
- dynamic programming
- average reward
- sparse linear systems
- learning algorithm
- machine learning
- multi objective
- special case
- training set
- optimal solution
- objective function
- data mining