Mode-matching control policies for multi-mode Markov decision processes.
Zhiyuan RenBruce H. KroghPublished in: ACC (2001)
Keyphrases
- markov decision processes
- control policies
- optimal policy
- action space
- finite horizon
- reinforcement learning
- state space
- reward function
- finite state
- dynamic programming
- continuous state
- transition matrices
- policy iteration
- decision problems
- average cost
- average reward
- infinite horizon
- partially observable
- markov decision process
- initial state
- long run
- multistage
- sufficient conditions
- decision theoretic planning
- motion control
- control strategies
- data mining
- computational complexity
- learning algorithm