Description:

Professor Sean Cornelius, Network Science Institute, Northeastern University

Title: The fundamental advantages of temporal networks

Abstract: It is increasingly recognized that many natural and social systems are best described as“temporal” networks, in which the underlying links exist only intermittently. For example, in metabolicnetworks the links correspond to relatively brief chemical reactions, and in social networksfriendship links are inferred from face-to-face or digital communications of finite duration.By effectively “fragmenting” a network's structure across time, temporality has been shown to have profound and usually deleterious effects on many dynamical processes, for example slowing down synchronization and the diffusion of innovative information, impeding exploration and navigation, and raising barriers to accessibility. Considering the ubiquity of temporal networks in nature, we must therefore ask: are there any advantages of temporality? In this talk, I will argue that there is at least one crucial dynamical process in which temporal networks enjoy unambiguous advantages, namely control—the ability to drive a system to (or keep it in) a desired state. Specifically, I will show that compared to their time-invariant counterparts, temporal systems can reach controllability faster, while demanding orders of magnitude less control energy and enjoying significantlymore compact control trajectories. These findings echo other recent work that has shown that nonlinear dynamics can be an asset, rather than an obstacle, to controlling real systems.

Host: Adilson Motter