Functional Stability of Discounted Markov Decision Processes Using Economic MPC Dissipativity Theory.
Arash Bahari KordabadSebastien GrosPublished in: ECC (2022)
Keyphrases
- markov decision processes
- optimal policy
- finite state
- state space
- infinite horizon
- average reward
- finite horizon
- policy iteration
- dynamic programming
- planning under uncertainty
- reinforcement learning
- reinforcement learning algorithms
- average cost
- action space
- decision theoretic planning
- model based reinforcement learning
- transition matrices
- markov decision process
- decision processes
- action sets
- discounted reward
- reachability analysis
- total reward
- partially observable
- decision diagrams
- state abstraction
- state and action spaces
- discount factor