Reinforcement Learning for Constrained Markov Decision Processes.
Ather GattamiQinbo BaiVaneet AggarwalPublished in: AISTATS (2021)
Keyphrases
- markov decision processes
- reinforcement learning
- reinforcement learning algorithms
- state space
- optimal policy
- policy iteration
- finite state
- model based reinforcement learning
- state and action spaces
- action space
- dynamic programming
- partially observable
- planning under uncertainty
- average reward
- markov decision process
- stochastic games
- state abstraction
- transition matrices
- learning algorithm
- control problems
- decision processes
- function approximation
- finite horizon
- average cost
- reward function
- factored mdps
- continuous state
- supervised learning
- action sets
- hierarchical reinforcement learning
- decision theoretic planning
- optimal control
- sufficient conditions
- markov chain
- single agent
- machine learning