When is Agnostic Reinforcement Learning Statistically Tractable?
Zeyu JiaGene LiAlexander RakhlinAyush SekhariNathan SrebroPublished in: CoRR (2023)
Keyphrases
- reinforcement learning
- function approximation
- np complete
- state space
- optimal policy
- reinforcement learning algorithms
- model free
- control problems
- computational complexity
- markov decision processes
- learning algorithm
- np hard
- action selection
- temporal difference learning
- computational problems
- markov decision process
- stochastic approximation
- reinforcement learning methods
- database
- temporal difference
- optimal control
- least squares
- multi agent
- neural network
- real time