Essentially Sharp Estimates on the Entropy Regularization Error in Discrete Discounted Markov Decision Processes.
Johannes MüllerSemih CayciPublished in: CoRR (2024)
Keyphrases
- markov decision processes
- state space
- optimal policy
- dynamic programming
- finite state
- finite horizon
- transition matrices
- reinforcement learning
- average reward
- infinite horizon
- reachability analysis
- policy iteration
- continuous state spaces
- decision processes
- average cost
- markov decision process
- action sets
- risk sensitive
- model based reinforcement learning
- reinforcement learning algorithms
- planning under uncertainty
- decision problems
- decision theoretic planning
- factored mdps
- partially observable
- state and action spaces
- partially observable markov decision processes
- discount factor
- interval estimation
- discounted reward
- total reward
- continuous state
- action space
- search space