Nearly Minimax Optimal Reinforcement Learning for Linear Markov Decision Processes.
Jiafan HeHeyang ZhaoDongruo ZhouQuanquan GuPublished in: CoRR (2022)
Keyphrases
- evaluation function
- markov decision processes
- reinforcement learning
- dynamic programming
- average cost
- optimal policy
- optimality criterion
- average reward
- reinforcement learning algorithms
- action sets
- finite horizon
- state space
- finite state
- policy iteration
- total reward
- markov decision process
- action space
- partially observable
- transition matrices
- model based reinforcement learning
- state and action spaces
- decision theoretic planning
- policy iteration algorithm
- discounted reward
- factored mdps
- optimal control
- stationary policies
- control policy
- approximate dynamic programming
- stochastic games
- infinite horizon
- state abstraction
- policy evaluation
- control policies
- planning under uncertainty
- function approximators
- model free
- function approximation
- machine learning
- optimal solution
- reachability analysis
- reward function