Optimistic PAC Reinforcement Learning: the Instance-Dependent View.
Andrea TirinzoniAymen Al MarjaniEmilie KaufmannPublished in: ALT (2023)
Keyphrases
- reinforcement learning
- multiple views
- function approximation
- neural network
- statistical queries
- learning algorithm
- multi agent reinforcement learning
- efficient learning
- model free
- state space
- dynamic programming
- optimal policy
- multi class
- temporal difference
- real robot
- temporal difference learning
- multi agent
- policy search