The statistical physics of real-world networks

Abstract

In the last 15 years, statistical physics has been a very successful framework to model complex networks. On the theoretical side, this approach has brought novel insights into a variety of physical phenomena, such as self-organisation, scale invariance, emergence of mixed distributions and ensemble non-equivalence, that display unconventional features on heterogeneous networks. At the same time, thanks to their deep connection with information theory, statistical physics and the principle of maximum entropy have led to the definition of null models for networks reproducing some features of real-world systems, but otherwise as random as possible. We review here the statistical physics approach and the various null models for complex networks, focusing in particular on the analytic frameworks reproducing the local network features. We then show how these models have been used to detect statistically significant and predictive structural patterns in real-world networks, as well as to reconstruct the network structure in case of incomplete information. We further survey the statistical physics models that reproduce more complex, semi-local network features using Markov chain Monte Carlo sampling, as well as the models of generalised network structures such as multiplex networks, interacting networks and simplicial complexes. arXiv:1810.05095v2 [physics.soc-ph]

Attack robustness and stability of generalized k-cores

“…. Exponential networks are common in many real-world networks [2,24,27] and they have average degree - which is approximately equal to the parameter σ for large σ. Figure 5 shows the behavior of n k (p) and l k (p) as functions of occupation fraction p for exponential networks with σ=80 under RA, LA, and TA with γ=1.…”

Scale-resolved analysis of brain functional connectivity networks with spectral entropy

“…Here we extend previous random graph models [43][44][45][46] to networks with real-valued links distributed over a connectivity backbone modeled by the degree and strength sequence. These local variables are the optimal trade-off to shape the irreducible and unavoidable complexity needed to accommodate the heterogeneous structure of real networks.…”

Minimum Entropy Stochastic Block Models Neglect Edge Distribution Heterogeneity

“…It is called the microcanonical ensemble (a vocabulary borrowed to statistical physics [7]) and it can be refined to impose that all graphs are simple, undirected (in which case M must be symmetric) and to allow or not self loops. In this paper we will consider multigraphs with self loops, because they allow for simpler computations.…”

“…To perform the second maximization, this method assumes that all graphs are generated with the same probability and it thus searches a partition of minimal entropy, in the sense that the cardinal of its microcanonical ensemble (i.e. the number of graphs the corresponding SBM can theoretically generate [7]) is minimal, which is equivalent to maximizing its likelihood [8]. In this paper, we show that even when the number and the size of the communities are fixed, the node partition which corresponds to the sharper communities is not always the one with the lower entropy.…”