Event details from: Renormalization of complex networks in latent space

Date
24 October 2024
Time
11:00 – 12:00
Room
Sweelinck

Title

Renormalization of complex networks in latent space

Abstract

The renormalization group in physics provides a rigorous approach to analyzing systems across different length scales. However, applying this method to complex networks is challenging due to the small-world property, which introduces correlations across coexisting scales. To address this, we have developed renormalization methods within the framework of network geometry, enabling the exploration of both weighted and unweighted real networks at different resolutions. This technique, based on progressively coarse-grained and rescaled network maps in hyperbolic space, has revealed multiscale self-similarity as a common symmetry in real networks. Additionally, this finding explains the self-similarity observed in anatomical multiscale reconstructions of human brain connectomes. Self-similarity is also evident in the growth patterns of certain networked systems, whose evolution can be effectively modeled by a reverse renormalization process. Practical applications of the self-similar multiscale unfolding of real networks include producing scaled replicas, which can be used in a variety of downstream tasks, such as studying processes where network size is relevant.

More about Mariangeles Serrano

Mariangeles Serrano is ICREA Research Professor at the University of Barcelona (UB). Professor Serrano is a distinguished expert in the fields of complex systems and network science, with a research focus on the hidden geometries and multiscale structures that shape complex networks in various domains, including biological, economic, and socio-technological systems.Her research provides new insights into the mechanisms driving the evolution and adaptation of these systems across different scales.

If you wish to attend this seminar online, please send an email to f.a.nobregasantos@uva.nl to receive the zoom-link.