Spectral Embedding of Weighted Graphs

Ian Gallagher, A. Jones, Anna Bertiger, Carey E. Priebe, Patrick Rubin‐DelanchyFigshare 2023 · DOI · OpenAlex

Abstract

When analyzing weighted networks using spectral embedding, a judicious transformation of the edge weights may produce better results. To formalize this idea, we consider the asymptotic behavior of spectral embedding for different edge-weight representations, under a generic low rank model. We measure the quality of different embeddings—which can be on entirely different scales—by how easy it is to distinguish communities, in an information-theoretical sense. For common types of weighted graphs, such as count networks or <i>p</i>-value networks, we find that transformations such as tempering or thresholding can be highly beneficial, both in theory and in practice. Supplementary materials for this article are available online.
BibTeX
@inproceedings{gallagher2023spectral,
  author = {Ian Gallagher and A. Jones and Anna Bertiger and Carey E. Priebe and Patrick Rubin‐Delanchy},
  title = {{Spectral Embedding of Weighted Graphs}},
  booktitle = {Figshare},
  year = {2023},
  doi = {10.6084/m9.figshare.23557217},
}

Loading related content...

Figure