Efficient Policy Iteration for Periodic Markov Decision Processes.
Takayuki OsogamiRudy RaymondPublished in: ECAI (2014)
Keyphrases
- markov decision processes
- policy iteration
- optimal policy
- finite state
- reinforcement learning
- transition matrices
- sample path
- state space
- factored mdps
- model free
- fixed point
- partially observable
- finite horizon
- dynamic programming
- decision processes
- planning under uncertainty
- least squares
- average reward
- average cost
- action space
- stochastic games
- policy evaluation
- markov decision process
- approximate dynamic programming
- reinforcement learning algorithms
- infinite horizon
- data mining
- discounted reward
- machine learning
- partially observable markov decision processes
- temporal difference
- convergence rate
- linear programming
- actor critic
- dynamical systems
- markov games
- policy iteration algorithm