A Reinforcement Learning Environment for Cooperative Multi-Agent Games: Enhancing Negotiation Skills
José Aleixo Cruz · Open Repository of the University of Porto (University of Porto) · 2019
In artificial intelligence, an agent is an autonomous program that interacts with its environment over time and decides its actions according to a given objective function.In a multi-agent system (MAS), multiple agents interact with each other.A MAS is decentralized, meaning that no agent controls other agents.Certain scenarios are inherently decentralized because the entities that inhabit it may need to act independently, with different or conflicting goals and information.Such situations can be naturally modeled as a MAS.José Aleixo Cruz v vi "A man provided with paper, pencil, and rubber, and subject to strict discipline, is in effect a universal machine."