Denumerable Undiscounted Semi-Markov Decision Processes with Unbounded Rewards.
Awi FedergruenPaul J. SchweitzerHenk C. TijmsPublished in: Math. Oper. Res. (1983)
Keyphrases
- semi markov decision processes
- markov decision processes
- average reward
- markov decision chains
- average cost
- reinforcement learning
- policy iteration
- finite state
- total reward
- optimal policy
- stochastic games
- state space
- infinite horizon
- reward function
- dynamic programming
- partially observable
- reinforcement learning algorithms
- decision processes
- markov decision problems
- markov decision process
- model free
- machine learning
- computational complexity
- action space
- long run
- markov chain