Optimality of Quasi-Open-Loop Policies for Discounted Semi-Markov Decision Processes.
Daniel AdelmanAngelo J. ManciniPublished in: Math. Oper. Res. (2016)
Keyphrases
- average reward
- open loop
- semi markov decision processes
- optimal policy
- closed loop
- markov decision processes
- average cost
- long run
- control system
- feedback control
- decision problems
- reinforcement learning
- dynamic programming
- finite state
- reward function
- inverted pendulum
- infinite horizon
- state space
- partially observable markov decision processes
- markov decision process
- policy iteration
- stability analysis
- control law
- initial state
- sufficient conditions
- optimal control
- markov chain
- control scheme
- partially observable
- machine learning
- fuzzy control
- stationary policies