Markovianity and mesoscopic structures on networks

Tuesday, November 12, 2019

2.30 pm

ISI seminar room 1st floor

Dr. Mauro Faccin, ICTEAM, Université Catholique de Louvain, Belgium

Complex systems often represent a challenge to the researcher that try  to understand their key components.
Community or block detection have proven to be an effective approach to  the simplification of complex systems when described as networks.
In this work we consider a dynamical system that evolve on a network;  the dynamics and the topology are both fundamental building blocks of  the systems and need to be considered together when analysing the system.
We use tools from the realm of Information Theory to characterize the  interaction of the dynamics and topology and to propose a new approach  to network partitioning that preserve their nature and mutual interaction.
An analogy with popular approaches to network partitioning is delineated  and few didactic and real world examples are proposed to support our  arguments.