On the adaptive control of a class of partially observed Markov decision processes.
Shun-Pin HsuDong-Ming ChuangAri ArapostathisPublished in: ACC (2009)
Keyphrases
- markov decision processes
- adaptive control
- partially observed
- reinforcement learning
- state space
- finite state
- optimal policy
- nonlinear systems
- transition matrices
- decision theoretic planning
- policy iteration
- control method
- factored mdps
- dynamic programming
- expected reward
- dynamic environments
- planning under uncertainty
- model based reinforcement learning
- control law
- reachability analysis
- control problems
- average reward
- finite horizon
- partially observable
- state and action spaces
- machine learning
- action space
- action sets
- infinite horizon
- function approximation
- average cost
- markov decision problems
- reward function
- mobile robot
- sufficient conditions
- finite number