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

Download:

Current browse context:

cond-mat

References & Citations

Bookmark

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

Condensed Matter > Disordered Systems and Neural Networks

Title: Systematic identification of statistically significant network measures

Abstract: We present a novel graph embedding space (i.e., a set of measures on graphs) for performing statistical analyses of networks. Key improvements over existing approaches include discovery of "motif-hubs" (multiple overlapping significant subgraphs), computational efficiency relative to subgraph census, and flexibility (the method is easily generalizable to weighted and signed graphs). The embedding space is based on {\it scalars}, functionals of the adjacency matrix representing the network. {\it Scalars} are global, involving all nodes; although they can be related to subgraph enumeration, there is not a one-to-one mapping between scalars and subgraphs. Improvements in network randomization and significance testing--we learn the distribution rather than assuming gaussianity--are also presented. The resulting algorithm establishes a systematic approach to the identification of the most significant scalars and suggests machine-learning techniques for network classification.
Comments: 19 pages, 12 figures
Subjects: Disordered Systems and Neural Networks (cond-mat.dis-nn); Statistical Mechanics (cond-mat.stat-mech); Quantitative Biology (q-bio)
Journal reference: Phys. Rev. E 71, 016110 (2005)
DOI: 10.1103/PhysRevE.71.016110
Cite as: arXiv:cond-mat/0306610 [cond-mat.dis-nn]
  (or arXiv:cond-mat/0306610v3 [cond-mat.dis-nn] for this version)

Submission history

From: Etay Ziv [view email]
[v1] Tue, 24 Jun 2003 17:48:55 GMT (291kb)
[v2] Wed, 25 Jun 2003 02:04:20 GMT (290kb)
[v3] Thu, 27 Jan 2005 22:01:36 GMT (38kb)

Link back to: arXiv, form interface, contact.