Abstract:A large variety of dynamical processes that take place on networks can be expressed in terms of the spectral properties of some linear operator which reflects how the dynamical rules depend on the network topology. Often such spectral features are theoretically obtained by considering only local node properties, such as degree distributions. Many networks, however, possess large-scale modular structures that can drastically influence their spectral characteristics, and which are neglected in such simplified descriptions. Here we obtain in a unified fashion the spectrum of a large family of operators, including the adjacency, Laplacian and normalized Laplacian matrices, for networks with generic modular structure, in the limit of large degrees. We focus on the conditions necessary for the merging of the isolated eigenvalues with the continuous band of the spectrum, after which the planted modular structure can no longer be easily detected by spectral methods. This is a crucial transition point which determines when a modular structure is strong enough to affect a given dynamical process. We show that this transition happens in general at different points for the different matrices, and hence the detectability threshold can vary significantly depending on the operator chosen. Equivalently, the sensitivity to the modular structure of the different dynamical processes associated with each matrix will be different, given the same large-scale structure present in the network. Furthermore, we show that, with the exception of the Laplacian matrix, the different transitions coalesce into the same point for the special case where the modules are homogeneous, but separate otherwise.

Joins of normal matrices, their spectrum, and applications

The spectrum of coupled random matrices

Spectrum of a class of matrices and its applications

Joint spectrum, group representations, and Julia set

Even spheres as joint spectra of matrix models

Tensor join of hypergraphs and its spectra

The Spectrum of the Singular Values of Z-Shaped Graph Matrices

On the spectrum of generalized H-join operation constrained by indexing maps -- I

New constructions of non-regular cospectral graphs

Eigenvalue Spectra of Modular Networks

The joint translation spectrum and Manhattan manifolds

Joint spectrum and large deviation principle for random matrix products

Convergent spectral inclusion sets for banded matrices

Spectra of eccentricity matrix of $H$-join of graphs

New matrices for spectral hypergraph theory, II

Polynomials on cyclic monotone elements with applications to random matrices with discrete spectrum

The generalized distance spectra of the -join of graphs, II

On the Representation of the Joint Spectrum for a Commutingn-Tuple of Non-Normal Operators

On Spectral Inclusion Sets and Computing the Spectra and Pseudospectra of Bounded Linear Operators

Invariant synchrony subspaces of sets of matrices

On the structure of join tensors with applications to tensor eigenvalue problems