#Cooperative Multi-Agent Reinforcement Learning(MARL)

Multi-agent reinforcement learning is an important part of swarm intelligence. A group of agents are able to achieve complex behaviours through cooperation. 

• research targets: self learning methods of cooperative strategies
• core techniques: deep reinforcement learning
• research directions: training pattern/ representaion learning/ exploration mechanisms/ cooperative emergent mechanisms
• applications: gaming ai

Academic Publications:

TCYB 2020 (IF: 10.387) / AAMAS 2020 / IJCNN 2019

 Future Directions

• Theoretical study: Deep reinforcement learning based cooperative group robots strtegy study.
• Realworld platform: 5G based multiple robots experiment platform.