Non-Asymptotic Analysis for Single-Loop (Natural) Actor-Critic with Compatible Function Approximation.
Yudan WangYue WangYi ZhouShaofeng ZouPublished in: CoRR (2024)
Keyphrases
- function approximation
- natural actor critic
- asymptotic analysis
- reinforcement learning
- reinforcement learning problems
- robot arm
- fluid model
- radial basis function
- temporal difference
- learning tasks
- function approximators
- model free
- reinforcement learning algorithms
- artificial neural networks
- reinforcement learning methods
- actor critic