Multi-Agent Reinforcement Learning: A Review of Challenges and Applications

In this review, we present an analysis of the most used multi-agent reinforcement learning algorithms. Starting with single-agent algorithms, focus on critical issues that must be taken into account in their extension to scenarios. The analyzed algorithms were grouped according features. We a detailed taxonomy main approaches proposed literature, focusing related mathematical models. For each algorithm, describe possible application fields, while pointing out its pros and cons. described are compared terms important characteristics for applications—namely, nonstationarity, scalability, observability. also common benchmark environments evaluate performances considered methods.