FlowPG: Action-constrained Policy Gradient with Normalizing Flows.
Janaka Chathuranga BrahmanageJiajing LingAkshat KumarPublished in: CoRR (2024)
Keyphrases
- policy gradient
- state action
- parametric optimization
- reinforcement learning
- gradient method
- optimal control
- actor critic
- reinforcement learning algorithms
- model free reinforcement learning
- function approximation
- variance reduction
- action space
- evaluation function
- policy search
- reinforcement learning methods
- function approximators
- approximation methods
- stochastic games
- action selection
- supervised learning
- markov decision process
- initial state
- dynamic programming
- computational complexity
- single agent
- multi agent
- neural network