Modeling and Control in Social Dynamics

Decision-Making and Leadership in Noisy Networks

Naomi Leonard

Princeton University

I will present recent results that make rigorous connections between graph measures and performance measures in network models of information assimilation and decision-making in uncertain environments. For decision-making between alternatives, I will show how a node’s speed-accuracy tradeoff is determined by its information centrality. I will then describe how information centrality extends to a notion of “joint centrality”, which distinguishes sets of leaders that minimize error in a network’s consensus decision about a noisy external signal. This is joint work with Vaibhav Srivastava and Katherine Fitch.