ACE: Cooperative Multi-agent Q-learning with Bidirectional Action-Dependency.
Chuming LiJie LiuYinmin ZhangYuhong WeiYazhe NiuYaodong YangYu LiuWanli OuyangPublished in: CoRR (2022)
Keyphrases
- cooperative multi agent
- reinforcement learning
- action selection
- state action
- cooperative
- learning algorithm
- state space
- function approximation
- stochastic approximation
- logic programming
- optimal policy
- markov decision processes
- action space
- multi agent
- machine learning
- initial state
- reasoning about actions
- decision making
- reinforcement learning methods
- agent learns