Natural Policy Gradient and Actor Critic Methods for Constrained Multi-Task Reinforcement Learning.
Sihan ZengThinh T. DoanJustin RombergPublished in: CoRR (2024)
Keyphrases
- policy gradient
- actor critic
- reinforcement learning
- multi task
- gradient method
- reinforcement learning algorithms
- function approximation
- model free
- control problems
- temporal difference
- optimal control
- metric learning
- reinforcement learning methods
- evaluation function
- learning tasks
- markov decision processes
- optimal policy
- neuro fuzzy
- machine learning
- optimization methods
- approximation methods
- average reward
- markov chain
- rl algorithms
- state space
- approximate dynamic programming
- learning process
- natural actor critic