Towards Tight Bounds on the Sample Complexity of Average-reward MDPs.
Yujia JinAaron SidfordPublished in: ICML (2021)
Keyphrases
- tight bounds
- average reward
- markov decision processes
- upper bound
- long run
- semi markov decision processes
- optimal policy
- reinforcement learning
- policy iteration
- discounted reward
- stochastic games
- optimality criterion
- state and action spaces
- dynamic programming
- model free
- total reward
- state space
- state action
- markov chain
- finite state
- decision theoretic planning
- reinforcement learning algorithms
- hierarchical reinforcement learning
- decision processes
- planning under uncertainty
- discount factor
- lower bound
- factored mdps
- reward function
- infinite horizon
- markov decision process