Multi-Scale Analysis on Complex Networks using Hermitian Graph Wavelets

We construct and study a class of spectral graph wavelets by analogy with Hermitian wavelets on the real line. We provide a localization result that significantly improves upon those previously available, enabling application to highly non-sparse, even complete, weighted graphs. We then define a new measure of importance of a node within a network called the Maximum Diffusion Time, and conclude by establishing an equivalence between the maximum diffusion time and information centrality, thus suggesting applications to quantifying hierarchical and distributed leadership structures in groups of interacting agents.