Adaptive control for discrete-time Markov processes with unbounded costs: Discounted criterion.
Evgueni GordienkoJ. Adolfo Minjárez-SosaPublished in: Kybernetika (1998)
Keyphrases
- markov processes
- adaptive control
- markov chain
- nonlinear systems
- markov process
- control method
- stochastic processes
- feedback control
- continuous time markov chains
- average cost
- random fields
- markov decision processes
- reinforcement learning
- non stationary
- control law
- dynamic environments
- infinite horizon
- dynamic programming
- least squares
- average reward
- optimal policy
- control parameters
- finite state
- steady state
- markov random field
- state space
- pairwise
- bayesian networks