Team-based Learning and Leadership Effect in Multi-agent System of Mixed Personalities
Aleksandar Stefanovski, P. Bock · Procedia Computer Science · 2013
This paper explores the effects of team-based learning for multi-agent system when mixed personalities are put together. Team- based learning is a policy for statistical learning systems that has been tested with teams of Collective Learning Automatons (CLA) to address a problem of joint learning. Each CLA is assigned a Five-Factor model of personality that affects its learning and decision-making. When teamed up, CLAs with some personalities show improved performance when playing the simple game of Nim.