English

Compatibility, embedding and regularization of non-local random walks on graphs

Numerical Analysis 2022-02-10 v2 Numerical Analysis

Abstract

Several variants of the graph Laplacian have been introduced to model non-local diffusion processes, which allow a random walker to {\textquotedblleft jump\textquotedblright} to non-neighborhood nodes, most notably the transformed path graph Laplacians and the fractional graph Laplacian. From a rigorous point of view, this new dynamics is made possible by having replaced the original graph GG with a weighted complete graph GG' on the same node-set, that depends on GG and wherein the presence of new edges allows a direct passage between nodes that were not neighbors in GG. We show that, in general, the graph GG' is not compatible with the dynamics characterizing the original model graph GG: the random walks on GG' subjected to move on the edges of GG are not stochastically equivalent, in the wide sense, to the random walks on GG. From a purely analytical point of view, the incompatibility of GG' with GG means that the normalized graph G^\hat{G} can not be embedded into the normalized graph G^\hat{G}'. Eventually, we provide a regularization method to guarantee such compatibility and preserving at the same time all the nice properties granted by GG'.

Keywords

Cite

@article{arxiv.2101.00425,
  title  = {Compatibility, embedding and regularization of non-local random walks on graphs},
  author = {Davide Bianchi and Marco Donatelli and Fabio Durastante and Mariarosa Mazza},
  journal= {arXiv preprint arXiv:2101.00425},
  year   = {2022}
}
R2 v1 2026-06-23T21:42:11.893Z