Learning and Planning in Average-Reward Markov Decision Processes.
Yi WanAbhishek NaikRichard S. SuttonPublished in: CoRR (2020)
Keyphrases
- markov decision processes
- average reward
- reinforcement learning
- stochastic games
- partially observable
- optimal policy
- policy iteration
- state space
- semi markov decision processes
- decision theoretic planning
- finite state
- actor critic
- state action
- learning algorithm
- factored mdps
- planning under uncertainty
- real time dynamic programming
- dynamic programming
- average cost
- partially observable markov decision processes
- discounted reward
- action space
- hierarchical reinforcement learning
- decision problems
- rl algorithms
- discount factor
- reinforcement learning algorithms
- infinite horizon
- model free
- action selection
- reward function
- radial basis function
- planning problems
- policy gradient
- policy evaluation
- dynamical systems
- heuristic search
- state and action spaces
- learning agent
- least squares