Computing a ranking network with confidence bounds from a graph-based Beta random field

Hsieh Fushing, Michael P. McAssey, Brenda Mccowan

Research output: Contribution to journalArticle

10 Citations (Scopus)

Abstract

We address two largely overlooked, fundamental issues in computing a ranking hierarchy within a society: which information in the network is relevant, and what effect chance has on the hierarchy. To properly account for uncertainty from limited data, we construct a random field in a matrix form having entry-wise posterior Beta distributions based on a graph of pairwise conflict outcomes. To evaluate relevant network information using information transitivity, another random matrix of synthesized transitive dominance odds is computed collectively along observed dominance paths. These two matrices are coupled together to fuse both direct and indirect dominance information. An ensemble of realizations of this fused random matrix facilitates an ensemble of optimal ranking networks by means of simulated annealing. Conditional statistical inferences regarding network features are derived, manifesting the effect of uncertainty. Our computational approach is suitable for large graphs of pairwise conflict outcomes, and can accommodate tremendous data heterogeneity-a typical feature in such studies. We also demonstrate the infeasibility of the classical maximum-likelihood approach, and expose the mechanistic flaws that stem from completely ignoring relevant information residing in the graph. We analyse two real datasets of decisive conflict outcomes, the first involving college football teams, and the second involving an adult rhesus macaque society in captivity.

Original languageEnglish (US)
Pages (from-to)3590-3612
Number of pages23
JournalProceedings of the Royal Society A: Mathematical, Physical and Engineering Sciences
Volume467
Issue number2136
DOIs
StatePublished - Dec 8 2011

Fingerprint

Confidence Bounds
ranking
Random Field
confidence
Ranking
Computing
Graph in graph theory
Random Matrices
hierarchies
Pairwise
Ensemble
Conditional Inference
Electric fuses
Uncertainty
Simulated annealing
Infeasibility
Odds
Beta distribution
fuses
Maximum likelihood

Keywords

  • Beta random field
  • Information transitivity
  • Nonlinear ranking hierarchy
  • Paired comparison
  • Rhesus macaque

ASJC Scopus subject areas

  • Mathematics(all)
  • Engineering(all)
  • Physics and Astronomy(all)

Cite this

Computing a ranking network with confidence bounds from a graph-based Beta random field. / Fushing, Hsieh; McAssey, Michael P.; Mccowan, Brenda.

In: Proceedings of the Royal Society A: Mathematical, Physical and Engineering Sciences, Vol. 467, No. 2136, 08.12.2011, p. 3590-3612.

Research output: Contribution to journalArticle

@article{0bba12d6cbc94e4fad93b970a149c2ac,
title = "Computing a ranking network with confidence bounds from a graph-based Beta random field",
abstract = "We address two largely overlooked, fundamental issues in computing a ranking hierarchy within a society: which information in the network is relevant, and what effect chance has on the hierarchy. To properly account for uncertainty from limited data, we construct a random field in a matrix form having entry-wise posterior Beta distributions based on a graph of pairwise conflict outcomes. To evaluate relevant network information using information transitivity, another random matrix of synthesized transitive dominance odds is computed collectively along observed dominance paths. These two matrices are coupled together to fuse both direct and indirect dominance information. An ensemble of realizations of this fused random matrix facilitates an ensemble of optimal ranking networks by means of simulated annealing. Conditional statistical inferences regarding network features are derived, manifesting the effect of uncertainty. Our computational approach is suitable for large graphs of pairwise conflict outcomes, and can accommodate tremendous data heterogeneity-a typical feature in such studies. We also demonstrate the infeasibility of the classical maximum-likelihood approach, and expose the mechanistic flaws that stem from completely ignoring relevant information residing in the graph. We analyse two real datasets of decisive conflict outcomes, the first involving college football teams, and the second involving an adult rhesus macaque society in captivity.",
keywords = "Beta random field, Information transitivity, Nonlinear ranking hierarchy, Paired comparison, Rhesus macaque",
author = "Hsieh Fushing and McAssey, {Michael P.} and Brenda Mccowan",
year = "2011",
month = "12",
day = "8",
doi = "10.1098/rspa.2011.0268",
language = "English (US)",
volume = "467",
pages = "3590--3612",
journal = "Proceedings of the Royal Society A: Mathematical, Physical and Engineering Sciences",
issn = "0080-4630",
publisher = "Royal Society of London",
number = "2136",

}

TY - JOUR

T1 - Computing a ranking network with confidence bounds from a graph-based Beta random field

AU - Fushing, Hsieh

AU - McAssey, Michael P.

AU - Mccowan, Brenda

PY - 2011/12/8

Y1 - 2011/12/8

N2 - We address two largely overlooked, fundamental issues in computing a ranking hierarchy within a society: which information in the network is relevant, and what effect chance has on the hierarchy. To properly account for uncertainty from limited data, we construct a random field in a matrix form having entry-wise posterior Beta distributions based on a graph of pairwise conflict outcomes. To evaluate relevant network information using information transitivity, another random matrix of synthesized transitive dominance odds is computed collectively along observed dominance paths. These two matrices are coupled together to fuse both direct and indirect dominance information. An ensemble of realizations of this fused random matrix facilitates an ensemble of optimal ranking networks by means of simulated annealing. Conditional statistical inferences regarding network features are derived, manifesting the effect of uncertainty. Our computational approach is suitable for large graphs of pairwise conflict outcomes, and can accommodate tremendous data heterogeneity-a typical feature in such studies. We also demonstrate the infeasibility of the classical maximum-likelihood approach, and expose the mechanistic flaws that stem from completely ignoring relevant information residing in the graph. We analyse two real datasets of decisive conflict outcomes, the first involving college football teams, and the second involving an adult rhesus macaque society in captivity.

AB - We address two largely overlooked, fundamental issues in computing a ranking hierarchy within a society: which information in the network is relevant, and what effect chance has on the hierarchy. To properly account for uncertainty from limited data, we construct a random field in a matrix form having entry-wise posterior Beta distributions based on a graph of pairwise conflict outcomes. To evaluate relevant network information using information transitivity, another random matrix of synthesized transitive dominance odds is computed collectively along observed dominance paths. These two matrices are coupled together to fuse both direct and indirect dominance information. An ensemble of realizations of this fused random matrix facilitates an ensemble of optimal ranking networks by means of simulated annealing. Conditional statistical inferences regarding network features are derived, manifesting the effect of uncertainty. Our computational approach is suitable for large graphs of pairwise conflict outcomes, and can accommodate tremendous data heterogeneity-a typical feature in such studies. We also demonstrate the infeasibility of the classical maximum-likelihood approach, and expose the mechanistic flaws that stem from completely ignoring relevant information residing in the graph. We analyse two real datasets of decisive conflict outcomes, the first involving college football teams, and the second involving an adult rhesus macaque society in captivity.

KW - Beta random field

KW - Information transitivity

KW - Nonlinear ranking hierarchy

KW - Paired comparison

KW - Rhesus macaque

UR - http://www.scopus.com/inward/record.url?scp=80755126944&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=80755126944&partnerID=8YFLogxK

U2 - 10.1098/rspa.2011.0268

DO - 10.1098/rspa.2011.0268

M3 - Article

AN - SCOPUS:80755126944

VL - 467

SP - 3590

EP - 3612

JO - Proceedings of the Royal Society A: Mathematical, Physical and Engineering Sciences

JF - Proceedings of the Royal Society A: Mathematical, Physical and Engineering Sciences

SN - 0080-4630

IS - 2136

ER -