Statistical Physics

15:40 - 16:50, Aula 9

Organizer: Eugenio Lippiello

Discussant: Salvatore Miccichè

StatPhys29: the IUPAP conference on Statistical Physics

Stefano Ruffo

Abstract: The Italian Society of Statistical Physics (SIFS) has proposed to host STATPHYS29, the International Union of Pure and Applied Physics (IUPAP) world conference on statistical physics, in Florence from July 13 to 18, 2025. The conference was presented in Tokyo at STATPHYS28 on August 11, 2023, and is currently in the planning stages. STATPHYS29 will cover a wide range of interdisciplinary topics and, for the first time, will include dedicated sessions on the statistical physics of machine learning.

Click here to view the abstract.

Statistical physics approach to brain activity

Lucilla De Arcangelis

Abstract: We give a short introduction to complex systems and present a number of examples from natural and biological processes. The standard statistical methods applied to characterize such processes is introduced for the special case of earthquakes, solar flares and neuronal signals. The comparison of results allows to highlight the different mechanisms driving activity in these three very different processes. In particular, the conditional probability analysis will be able to evidence the special role that the alpha rhythm plays in the rest activity of human brains.

Click here to view the abstract.

Information Core and Laplacian Renormalization Group for Complex Networks

Andrea Gabrielli, Tommaso Gili, Guido Caldarelli, Pablo Villegas

Abstract: We propose a novel diffusion-driven Laplacian Renormalization Group (LRG) scheme to identify relevant spatio-temporal scales in heterogeneous networks. Firstly, we present a heuristic real-space RG inspired by Migdal-Kadanoff prescriptions, addressing issues like small-world effects and decimation problems. This framework integrates fast diffusion modes progressively, defining coarsegrained macronodes and connections, and a renormalized Laplacian operator. We then introduce a more rigorous version of the diffusion-based RG, akin to Wilson’s k−space RG in statistical field theory. We apply LRG to various networks, demonstrating its ability to capture essential system properties and perform network reduction and change of resolution scale while connecting the behavior of the related entropic susceptibility with scale-invariant network properties.

Click here to view the abstract.