Safe Reinforcement Learning in Constrained Markov Decision Processes.
Akifumi WachiYanan SuiPublished in: CoRR (2020)
Keyphrases
- markov decision processes
- reinforcement learning
- reinforcement learning algorithms
- state space
- optimal policy
- finite state
- policy iteration
- action space
- model based reinforcement learning
- partially observable
- markov decision process
- average reward
- reachability analysis
- infinite horizon
- dynamic programming
- action sets
- state and action spaces
- finite horizon
- reward function
- transition matrices
- decision theoretic planning
- average cost
- stochastic games
- semi markov decision processes
- planning under uncertainty
- temporal difference
- state abstraction
- decentralized control
- factored mdps
- total reward
- stochastic shortest path
- approximate dynamic programming
- policy evaluation
- continuous state
- decision processes
- model free
- optimal control
- least squares
- multi agent