More Benefits of Being Distributional: Second-Order Bounds for Reinforcement Learning.
Kaiwen WangOwen OertellAlekh AgarwalNathan KallusWen SunPublished in: CoRR (2024)
Keyphrases
- reinforcement learning
- upper bound
- function approximation
- higher order
- upper and lower bounds
- state space
- optimal policy
- machine learning
- tight bounds
- model free
- lower bound
- learning process
- error bounds
- learning algorithm
- robotic control
- co occurrence
- dynamic programming
- markov decision processes
- learning classifier systems
- vc dimension
- genetic algorithm
- reinforcement learning algorithms
- robot control
- markov decision process
- real time