VOQL: Towards Optimal Regret in Model-free RL with Nonlinear Function Approximation.
Alekh AgarwalYujia JinTong ZhangPublished in: CoRR (2022)
Keyphrases
- model free
- function approximation
- reinforcement learning
- reinforcement learning algorithms
- temporal difference
- average reward
- radial basis function
- rl algorithms
- learning tasks
- policy iteration
- state space
- temporal difference learning
- dynamic programming
- temporal difference methods
- neural network
- function approximators
- learning process
- td learning
- policy evaluation
- control policy
- least squares
- reward function
- optimal policy
- collaborative filtering