The perfect team of the future will most likely consist of humans and agents(AI). To this purpose, we conducted experiments with human subjects to explore how they integrate their individual decisions into a group response, how much they rely on agents in their team, and how an agent's input affects their decision making process.
Second, we propose models investigating the dataset with multiple approaches employing Bayes rule, eigenvector centrality of human influence and expected agent performance, along with Prospect Theory applied to the aforementioned algorithms.
Finally, I will discuss our findings and the conclusions we draw from our results, in which we attempt to better understand and predict team behavior. We propose that successful incorporation of our findings in building human-agent teams will greatly aid the team decision making process and increase the performance of such teams.