Constrained Markov Decision Processes via Backward Value Functions.
Harsh SatijaPhilip AmortilaJoelle PineauPublished in: CoRR (2020)
Keyphrases
- markov decision processes
- optimal policy
- finite state
- transition matrices
- dynamic programming
- state space
- reachability analysis
- reinforcement learning
- decision theoretic planning
- finite horizon
- policy iteration
- decision diagrams
- reinforcement learning algorithms
- average cost
- model based reinforcement learning
- infinite horizon
- partially observable
- decision processes
- average reward
- state and action spaces
- risk sensitive
- model free
- search space
- factored mdps
- markov decision process
- stationary policies
- planning under uncertainty
- basis functions
- markov chain
- objective function