Behavior Analysis of Children Using Inverse Reinforcement Learning in Group Discussion

Abstract

In recent years, there has been a growing need to enhance the non-cognitive abilities of children in childcare settings. For this reason, it is necessary for childcare workers to teach children how to change their behavior to improve their non-cognitive abilities. However, in order to promote behavioral change, it is necessary to understand the situations and motivations under which children act. In this study, analyzes children’s behavior in group discussions using inverse reinforcement learning, which aims to search for reward functions that explain observed states and behaviors.