On Exchangeability in Network Models

Steffen L. Lauritzen; Alessandro Rinaldo; Kayvan Sadeghi

doi:10.18409/jas.v10i1.73

On Exchangeability in Network Models

Steffen L. Lauritzen, Alessandro Rinaldo, Kayvan Sadeghi

Department of Mathematical Sciences

2 Downloads (Pure)

Abstract

We derive representation theorems for exchangeable distributions on finite and infinite graphs using elementary arguments based on geometric and graph-theoretic concepts. Our results elucidate some of the key differences, and their implications, between statistical network models that are finitely exchangeable and models that define a consistent sequence of probability distributions on graphs of increasing size.

Original language	English
Journal	Journal of Algebraic Statistics
Volume	10
Issue number	1
Pages (from-to)	85-114
ISSN	1309-3452
DOIs	https://doi.org/10.18409/jas.v10i1.73
Publication status	Published - 12 Apr 2019

Access to Document

10.18409/jas.v10i1.73Licence: CC BY-NC

OA-On Exchangeability in Network ModelsFinal published version, 875 KBLicence: CC BY-NC

Cite this

@article{86159650156b4b8483580a26db5feed5,

title = "On Exchangeability in Network Models",

abstract = "We derive representation theorems for exchangeable distributions on finite and infinite graphs using elementary arguments based on geometric and graph-theoretic concepts. Our results elucidate some of the key differences, and their implications, between statistical network models that are finitely exchangeable and models that define a consistent sequence of probability distributions on graphs of increasing size. ",

author = "Lauritzen, {Steffen L.} and Alessandro Rinaldo and Kayvan Sadeghi",

year = "2019",

month = apr,

day = "12",

doi = "10.18409/jas.v10i1.73",

language = "English",

volume = "10",

pages = "85--114",

journal = "Journal of Algebraic Statistics",

issn = "1309-3452",

publisher = "Journal of Algebraic Statistics",

number = "1",

}

TY - JOUR

T1 - On Exchangeability in Network Models

AU - Lauritzen, Steffen L.

AU - Rinaldo, Alessandro

AU - Sadeghi, Kayvan

PY - 2019/4/12

Y1 - 2019/4/12

N2 - We derive representation theorems for exchangeable distributions on finite and infinite graphs using elementary arguments based on geometric and graph-theoretic concepts. Our results elucidate some of the key differences, and their implications, between statistical network models that are finitely exchangeable and models that define a consistent sequence of probability distributions on graphs of increasing size.

AB - We derive representation theorems for exchangeable distributions on finite and infinite graphs using elementary arguments based on geometric and graph-theoretic concepts. Our results elucidate some of the key differences, and their implications, between statistical network models that are finitely exchangeable and models that define a consistent sequence of probability distributions on graphs of increasing size.

U2 - 10.18409/jas.v10i1.73

DO - 10.18409/jas.v10i1.73

M3 - Journal article

SN - 1309-3452

VL - 10

SP - 85

EP - 114

JO - Journal of Algebraic Statistics

JF - Journal of Algebraic Statistics

IS - 1

ER -

On Exchangeability in Network Models

Abstract

Access to Document

Fingerprint

Cite this