RAMBO-RL: Robust Adversarial Model-Based Offline Reinforcement Learning.
Marc RigterBruno LacerdaNick HawesPublished in: CoRR (2022)
Keyphrases
- reinforcement learning
- model free
- multi agent
- reinforcement learning algorithms
- function approximation
- state space
- markov decision processes
- learning algorithm
- markov decision process
- temporal difference
- action space
- rl algorithms
- optimal policy
- approximate dynamic programming
- learning problems
- reinforcement learning methods
- action selection
- real time
- dynamic programming
- optimal control
- multi agent reinforcement learning
- direct policy search
- policy iteration
- transfer learning
- supervised learning
- neural network