Multi-Agent Reinforcement Learning for Exploring Collective Behavior

Humans and other organisms develop collective behaviors through interactions with diverse environments and various species. These behaviors are significant topics across multiple research fields, including evolutionary biology, behavioral ecology, and animal sociology. Unraveling the decision-making mechanisms of individuals in groups within cooperative and competitive contexts has captured the attention of many researchers but remains a complex challenge. This seminar will present research cases that employ multi-agent reinforcement learning, a machine learning technique, to investigate the decision-making processes underlying collective behavior. Through this approach, we aim to provide deeper insights into the dynamics and mechanisms that drive group behaviors in various biological systems.