Home > Research > Publications & Outputs > Modeling Network Populations via Graph Distances

Electronic data

  • Modeling Network Populations via Graph Distances

    Rights statement: This is an Accepted Manuscript of an article published by Taylor & Francis in Journal of the American Statistical Association on 6 May 2020, available online: https://www.tandfonline.com/doi/full/10.1080/01621459.2020.1763803

    Accepted author manuscript, 1.69 MB, PDF document

    Embargo ends: 6/05/21

    Available under license: CC BY-NC: Creative Commons Attribution-NonCommercial 4.0 International License

Links

Text available via DOI:

View graph of relations

Modeling Network Populations via Graph Distances

Research output: Contribution to journalJournal articlepeer-review

E-pub ahead of print

Standard

Modeling Network Populations via Graph Distances. / Lunagomez Coria, Simon; Olhede, Sofia Charlotta; Wolfe, Patrick.

In: Journal of the American Statistical Association, 06.05.2020.

Research output: Contribution to journalJournal articlepeer-review

Harvard

Lunagomez Coria, S, Olhede, SC & Wolfe, P 2020, 'Modeling Network Populations via Graph Distances', Journal of the American Statistical Association. https://doi.org/10.1080/01621459.2020.1763803

APA

Lunagomez Coria, S., Olhede, S. C., & Wolfe, P. (2020). Modeling Network Populations via Graph Distances. Journal of the American Statistical Association. https://doi.org/10.1080/01621459.2020.1763803

Vancouver

Lunagomez Coria S, Olhede SC, Wolfe P. Modeling Network Populations via Graph Distances. Journal of the American Statistical Association. 2020 May 6. https://doi.org/10.1080/01621459.2020.1763803

Author

Lunagomez Coria, Simon ; Olhede, Sofia Charlotta ; Wolfe, Patrick. / Modeling Network Populations via Graph Distances. In: Journal of the American Statistical Association. 2020.

Bibtex

@article{48619267d3484dfea90e3a9bc7bfd0d9,
title = "Modeling Network Populations via Graph Distances",
abstract = "This article introduces a new class of models for multiple networks. The core idea is to parametrize a distribution on labelled graphs in terms of a Fr{\'e}chet mean graph (which depends on a user-specified choice of metric or graph distance) and a parameter that controls the concentration of this distribution about its mean. Entropy is the natural parameter for such control, varying from a point mass concentrated on the Fr{\'e}chet mean itself to a uniform distribution over all graphs on a given vertex set. We provide a hierarchical Bayesian approach for exploiting this construction, along with straightforward strategies for sampling from the resultant posterior distribution. We conclude by demonstrating the efficacy of our approach via simulation studies and two multiple-network data analysis examples: one drawn from systems biology and the other from neuroscience.",
keywords = "Hierarchical Bayesian models, Graph metrics, Network variability, Object oriented data, Random graphs, Statistical network analysis",
author = "{Lunagomez Coria}, Simon and Olhede, {Sofia Charlotta} and Patrick Wolfe",
note = "This is an Accepted Manuscript of an article published by Taylor & Francis in Journal of the American Statistical Association on 6 May 2020, available online: https://www.tandfonline.com/doi/full/10.1080/01621459.2020.1763803 ",
year = "2020",
month = may,
day = "6",
doi = "10.1080/01621459.2020.1763803",
language = "English",
journal = "Journal of the American Statistical Association",
issn = "0162-1459",
publisher = "Taylor and Francis Ltd.",

}

RIS

TY - JOUR

T1 - Modeling Network Populations via Graph Distances

AU - Lunagomez Coria, Simon

AU - Olhede, Sofia Charlotta

AU - Wolfe, Patrick

N1 - This is an Accepted Manuscript of an article published by Taylor & Francis in Journal of the American Statistical Association on 6 May 2020, available online: https://www.tandfonline.com/doi/full/10.1080/01621459.2020.1763803

PY - 2020/5/6

Y1 - 2020/5/6

N2 - This article introduces a new class of models for multiple networks. The core idea is to parametrize a distribution on labelled graphs in terms of a Fréchet mean graph (which depends on a user-specified choice of metric or graph distance) and a parameter that controls the concentration of this distribution about its mean. Entropy is the natural parameter for such control, varying from a point mass concentrated on the Fréchet mean itself to a uniform distribution over all graphs on a given vertex set. We provide a hierarchical Bayesian approach for exploiting this construction, along with straightforward strategies for sampling from the resultant posterior distribution. We conclude by demonstrating the efficacy of our approach via simulation studies and two multiple-network data analysis examples: one drawn from systems biology and the other from neuroscience.

AB - This article introduces a new class of models for multiple networks. The core idea is to parametrize a distribution on labelled graphs in terms of a Fréchet mean graph (which depends on a user-specified choice of metric or graph distance) and a parameter that controls the concentration of this distribution about its mean. Entropy is the natural parameter for such control, varying from a point mass concentrated on the Fréchet mean itself to a uniform distribution over all graphs on a given vertex set. We provide a hierarchical Bayesian approach for exploiting this construction, along with straightforward strategies for sampling from the resultant posterior distribution. We conclude by demonstrating the efficacy of our approach via simulation studies and two multiple-network data analysis examples: one drawn from systems biology and the other from neuroscience.

KW - Hierarchical Bayesian models

KW - Graph metrics

KW - Network variability

KW - Object oriented data

KW - Random graphs

KW - Statistical network analysis

U2 - 10.1080/01621459.2020.1763803

DO - 10.1080/01621459.2020.1763803

M3 - Journal article

JO - Journal of the American Statistical Association

JF - Journal of the American Statistical Association

SN - 0162-1459

ER -