MAAIGTLGDec 8, 2023

A Review of Cooperation in Multi-agent Learning

arXiv:2312.05162v153 citationsh-index: 43
Originality Synthesis-oriented
AI Analysis

It addresses the problem of understanding and improving cooperation in multi-agent systems for researchers and practitioners, but is incremental as it is a review paper.

This paper provides an overview of fundamental concepts, problem settings, and algorithms in multi-agent learning, focusing on cooperation, and reviews recent progress and open challenges in the field.

Cooperation in multi-agent learning (MAL) is a topic at the intersection of numerous disciplines, including game theory, economics, social sciences, and evolutionary biology. Research in this area aims to understand both how agents can coordinate effectively when goals are aligned and how they may cooperate in settings where gains from working together are possible but possibilities for conflict abound. In this paper we provide an overview of the fundamental concepts, problem settings and algorithms of multi-agent learning. This encompasses reinforcement learning, multi-agent sequential decision-making, challenges associated with multi-agent cooperation, and a comprehensive review of recent progress, along with an evaluation of relevant metrics. Finally we discuss open challenges in the field with the aim of inspiring new avenues for research.

Foundations

The foundational work for this paper's niche, ranked by how specifically the neighbourhood builds on it — not by global fame.

Your Notes