Safe Reinforcement Learning for Constrained Markov Decision Processes with Stochastic Stopping Time.
Abhijit MazumdarRafal WisniewskiManuela-Luminita BujorianuPublished in: CoRR (2024)
Keyphrases
- markov decision processes
- reinforcement learning
- optimal policy
- policy iteration algorithm
- reinforcement learning algorithms
- continuous state spaces
- state space
- action space
- dynamic programming
- finite state
- policy iteration
- state and action spaces
- control policies
- transition matrices
- reward function
- model based reinforcement learning
- planning under uncertainty
- average reward
- partially observable
- markov decision process
- finite horizon
- state dependent
- action sets
- monte carlo
- average cost
- decision theoretic planning
- infinite horizon
- model free
- factored mdps
- function approximation
- decentralized control
- reachability analysis
- state abstraction
- control problems
- temporal difference
- multi agent
- approximate dynamic programming
- stochastic games
- optimal control