Abstract:Many natural and social systems develop complex networks, that are usually modelled as random graphs. The eigenvalue spectrum of these graphs provides information about their structural properties. While the semi-circle law is known to describe the spectral density of uncorrelated random graphs, much less is known about the eigenvalues of real-world graphs, describing such complex systems as the Internet, metabolic pathways, networks of power stations, scientific collaborations or movie actors, which are inherently correlated and usually very sparse. An important limitation in addressing the spectra of these systems is that the numerical determination of the spectra for systems with more than a few thousand nodes is prohibitively time and memory consuming. Making use of recent advances in algorithms for spectral characterization, here we develop new methods to determine the eigenvalues of networks comparable in size to real systems, obtaining several surprising results on the spectra of adjacency matrices corresponding to models of real-world graphs. We find that when the number of links grows as the number of nodes, the spectral density of uncorrelated random graphs does not converge to the semi-circle law. Furthermore, the spectral densities of real-world graphs have specific features depending on the details of the corresponding models. In particular, scale-free graphs develop a triangle-like spectral density with a power law tail, while small-world graphs have a complex spectral density function consisting of several sharp peaks. These and further results indicate that the spectra of correlated graphs represent a practical tool for graph classification and can provide useful insight into the relevant structural properties of real networks.

Isospectral Reductions of Dynamical Networks

Isospectral Reductions of Non-negative Matrices

Identifiability and Identification of Switching Dynamical Networks: A Data-Based Approach

Laplacian Spectra of a Class of Small-World Networks and Their Applications

Dimension reduction of dynamics on modular and heterogeneous directed networks

Spectral energy transfer on complex networks: a filtering approach

Spectral identification of networks with generalized diffusive coupling

Dimensionality reduction of networked systems with separable coupling-dynamics: theory and applications

Comprehensive spectral approach for community structure analysis on complex networks

Spectral Graph Analysis: A Unified Explanation and Modern Perspectives

Dimensionality Reduction in Discrete-Time Dynamical Systems

Spectra of "Real-World" Graphs: Beyond the Semi-Circle Law

Measuring hierarchically-organized interactions in dynamic networks through spectral entropy rates: theory, estimation, and illustrative application to physiological networks

Eigenvalue Spectra of Modular Networks

From graphs to signals and back: Identification of network structures using spectral analysis

Inferring the Graph of Networked Dynamical Systems under Partial Observability and Spatially Colored Noise

Minimal realization of the dynamical structure function and its application to network reconstruction

Isospectral graphs via inner symmetries

Sparse dynamic network reconstruction through L1-regularization of a Lyapunov equation

Model reduction to spectral submanifolds in piecewise smooth dynamical systems

Neural Network Spectral Robustness under Perturbations of the Underlying Graph