Non-Linear Reinforcement Learning in Large Action Spaces: Structural Conditions and Sample-efficiency of Posterior Sampling.
Alekh AgarwalTong ZhangPublished in: COLT (2022)
Keyphrases
- action space
- reinforcement learning
- state space
- markov decision processes
- continuous state
- state and action spaces
- real valued
- sample size
- control policies
- stochastic processes
- reinforcement learning methods
- function approximation
- action selection
- state action
- sufficient conditions
- reinforcement learning algorithms
- dynamic programming
- function approximators
- skill learning
- continuous state spaces
- reinforcement learning problems
- learning algorithm
- markov decision process
- optimal policy
- single agent
- dynamical systems
- computational complexity
- decision making
- machine learning