We gratefully acknowledge support from
the Simons Foundation and member institutions.
Full-text links:

Download:

Current browse context:

cs.SI

Change to browse by:

cs

References & Citations

DBLP - CS Bibliography

Bookmark

(what is this?)
CiteULike logo BibSonomy logo Mendeley logo del.icio.us logo Digg logo Reddit logo ScienceWISE logo

Computer Science > Social and Information Networks

Title: Graph space: using both geometric and probabilistic structure to evaluate statistical graph models

Abstract: Statistical graph models aim at modeling graphs as random realization among a set of possible graphs. One issue is to evaluate whether or not a graph is likely to have been generated by one particular model. In this paper we introduce the edit distance expected value (EDEV) and compare it with other methods such as entropy and distance to the barycenter. We show that contrary to them, EDEV is able to distinguish between graphs that have a typical structure with respect to a model, and those that do not. Finally we introduce a statistical hypothesis testing methodology based on this distance to evaluate the relevance of a candidate model with respect to an observed graph.
Subjects: Social and Information Networks (cs.SI)
Cite as: arXiv:2106.13587 [cs.SI]
  (or arXiv:2106.13587v1 [cs.SI] for this version)

Submission history

From: Louis Duvivier [view email]
[v1] Fri, 25 Jun 2021 12:38:50 GMT (1480kb)

Link back to: arXiv, form interface, contact.