Multi-Agent Actor-Critic for Mixed Cooperative-Competitive Environments.
Ryan LoweYi WuAviv TamarJean HarbPieter AbbeelIgor MordatchPublished in: NIPS (2017)
Keyphrases
- cooperative
- multi agent
- actor critic
- reinforcement learning
- multi agent systems
- optimal control
- policy gradient
- temporal difference
- approximate dynamic programming
- gradient method
- single agent
- function approximation
- dynamic environments
- game theory
- reinforcement learning algorithms
- markov decision processes
- policy iteration
- average reward
- neural network