Inexperienced RL Agents Can't Get It Right: Lower Bounds on Regret at Finite Sample Complexity.
Maia FraserVincent LétourneauPublished in: CoLLAs (2022)
Keyphrases
- lower bound
- finite sample
- worst case
- sample size
- upper bound
- statistical learning theory
- error bounds
- reinforcement learning
- vc dimension
- uniform convergence
- nearest neighbor
- np hard
- lower and upper bounds
- objective function
- upper and lower bounds
- generalization error
- state space
- parzen window
- support vector machine
- special case
- optimal solution