A statistical mechanics approach to inverse dynamical problems on sparse graphs
Luogo: Aula Newton
Relatore: Luca Dall'Asta
Note sul Relatore: Dipartimento di Scienza Applicata e Tecnologia, Politecnico di Torino
Abstract:
Who are the most influential individuals in a social network? What's the origin of an outbreak? Simple questions like these hide incredibly difficult theoretical and computational problems. I will present a statistical mechanics approach, based on the cavity method, to study inverse dynamical problems in the simplified framework provided by monotone dynamics on sparse graphs. I will review applications to viral marketing, to epidemic inference and network controllability and discuss potential and limitations of such methods.
