Efficient Strategy Iteration for Mean Payoff in Markov Decision Processes.
Jan KretínskýTobias MeggendorferPublished in: ATVA (2017)
Keyphrases
- markov decision processes
- state space
- optimal policy
- finite state
- reinforcement learning
- reinforcement learning algorithms
- transition matrices
- decision theoretic planning
- stochastic games
- risk sensitive
- policy iteration
- finite horizon
- action space
- planning under uncertainty
- model based reinforcement learning
- infinite horizon
- average cost
- factored mdps
- reachability analysis
- decision processes
- dynamic programming
- data mining
- action sets