Graphon-based sensitivity analysis of SIS epidemics

TL;DR

This paper employs graphon spectral properties to analyze the stability and noise sensitivity of SIS epidemic models on large networks, providing a new framework for understanding epidemic dynamics in complex systems.

Contribution

It introduces a novel graphon-based approach to quantify the sensitivity of SIS epidemics to noise, extending analysis from finite networks to large random graph models.

Findings

Derived a noise index based on adjacency eigenvalues for finite networks.

Provided an approximation of the noise index using graphon eigenvalues for large networks.

Numerical example demonstrating the theoretical results.

Abstract

In this work, we use the spectral properties of graphons to study stability and sensitivity to noise of deterministic SIS epidemics over large networks. We consider the presence of additive noise in a linearized SIS model and we derive a noise index to quantify the deviation from the disease-free state due to noise. For finite networks, we show that the index depends on the adjacency eigenvalues of its graph. We then assume that the graph is a random sample from a piecewise Lipschitz graphon with finite rank and, using the eigenvalues of the associated graphon operator, we find an approximation of the index that is tight when the network size goes to infinity. A numerical example is included to illustrate the results.