Dimensions of machine learning in design
Artificial Intelligence for Engineering Design, Analysis and Manufacturing
Autonomous Agents that Learn to Better Coordinate
Autonomous Agents and Multi-Agent Systems
Modeling collective learning in design
Artificial Intelligence for Engineering Design, Analysis and Manufacturing
Cooperative Multi-Agent Learning: The State of the Art
Autonomous Agents and Multi-Agent Systems
| Hi-index | 0.00 |
This paper describes the relationship between team performance and the modes of social learning and level of team familiarity, i.e. agents that have worked together before. A computational model is implemented under a set of typical social learning modes in a team environment. In this model, agents can learn from personal interaction with other agents and their tasks, and by observing interaction among other agents and the tasks. Agents in the team are considered domain experts, which means the task knowledge is pre-coded. Agents learn about each others' competence (i.e., who knows what), which leads to the formation of a team mental model. Agents that have team familiarity are expected to have developed each other's mental model to the extent facilitated by the available learning modes. Simulations are conducted with team familiarity and learning modes as parameters. Simulation results indicate that team performance is positively correlated with social learning and team familiarity. Implications of the findings on managing information exchange within teams are discussed.