Advances on Multi-agent Hierarchical Reinforcement Learning
Guochang Gu · The Border Economy and Culture · 2007
Reinforcement learning(RL) is an approach that an agent can learn its behaviors through trial-and-error interaction with a dynamic environment.It has been an important branch of machine learning for its self-learning and online learning capabilities.Multi-agent reinforcement learning(MARL) and hierarchical reinforcement learning(HRL) are two of the focuses in the domain recently.And multi-agent hierarchical reinforcement learning(MAHRL),by integrating MARL and HRL,is a new topic.The efforts on MARL,HRL,and MAHRL are achieved and the open challenges in MAHRL domain are stressed.