Accelerating Interval Iteration for Expected Rewards in Markov Decision Processes.
MohammadSadegh MohagheghiKhayyam SalehiPublished in: ICSOFT (2020)
Keyphrases
- markov decision processes
- total reward
- reinforcement learning
- optimal policy
- finite state
- state space
- dynamic programming
- transition matrices
- policy iteration
- planning under uncertainty
- reinforcement learning algorithms
- finite horizon
- factored mdps
- partially observable
- average cost
- decision theoretic planning
- average reward
- discounted reward
- stationary policies
- sequential decision making under uncertainty
- action space
- reward function
- reachability analysis
- model based reinforcement learning
- decision problems
- semi markov decision processes
- decentralized control
- markov decision process
- action sets
- markov chain
- search space
- objective function