Regensburg 2019 – wissenschaftliches Programm
SOE 12.12: Poster
Dienstag, 2. April 2019, 16:00–19:00, Poster A
Network reconstruction based on event timing in spreading processes — Forough Hassanibesheli1,2 and •Reik V. Donner1,3 — 1PIK Potsdam, Germany — 2Humboldt University, Potsdam, Germany — 3Magdeburg-Stendal University of Applied Sciences, Magdeburg, Germany
Spreading phenomena like opinion formation or disease propagation commonly follow the links of some underlying network structure. While the effect of network topology on spreading efficiency has already been vastly studied, we here address the inverse problem whether we can infer an unknown network structure from the timing of events at different nodes. For this purpose, we consider a simple model of nodes exhibiting two types of activity: spontaneous events that are generated via mutually independent Poisson processes, and triggered events that occur with a certain probability whenever one of the neighboring nodes exhibits any of these two kinds of events. Based on simulations of this model for different types of networks, we study the similarity between the timings of events at all pairs of nodes by means of event synchronization and event coincidence analysis as two wide-spread methods for studying simultaneity in event series. By taking strong mutual similarities as proxies for actual physical links, we demonstrate that both approaches lead to similar prediction accuracy. In general, sparser networks can be reconstructed more accurately than denser ones, especially in case of larger networks.