A note on discrete multivariate Markov random field models

H.L. Ip, K.Y. Karl Wu

Research output: Contribution to journalArticle

Abstract

Markov random field (MRF) is commonly used in modelling spatially dependent data. These models are often referred to as auto-models. While univariate auto-models have been extensively studied in the literature, discrete multivariate MRF has not attracted much attention. This paper attempts to fill the research gap by providing some results on the discrete multivariate MRF scheme, which forms the theoretical foundation to construct models for spatially dependent categorical data. The results presented in this paper allow the formulation of a novel auto-model and justify the validity of the recently proposed auto-multinomial model.
Original languageEnglish
Article number108588
Pages (from-to)1-6
Number of pages6
JournalStatistics and Probability Letters
Volume156
Early online date19 Aug 2019
DOIs
Publication statusPublished - Jan 2020

Fingerprint

Random Field
Dependent Data
Multinomial Model
Model
Nominal or categorical data
Justify
Univariate
Random field
Formulation
Modeling

Cite this

@article{84f3892428ba445b93ccc0e5d5e78157,
title = "A note on discrete multivariate Markov random field models",
abstract = "Markov random field (MRF) is commonly used in modelling spatially dependent data. These models are often referred to as auto-models. While univariate auto-models have been extensively studied in the literature, discrete multivariate MRF has not attracted much attention. This paper attempts to fill the research gap by providing some results on the discrete multivariate MRF scheme, which forms the theoretical foundation to construct models for spatially dependent categorical data. The results presented in this paper allow the formulation of a novel auto-model and justify the validity of the recently proposed auto-multinomial model.",
keywords = "Auto-models, Lattice system, MRF, Spatial correlation",
author = "H.L. Ip and Wu, {K.Y. Karl}",
year = "2020",
month = "1",
doi = "10.1016/j.spl.2019.108588",
language = "English",
volume = "156",
pages = "1--6",
journal = "Statistics and Probability Letters",
issn = "0167-7152",
publisher = "Elsevier",

}

A note on discrete multivariate Markov random field models. / Ip, H.L.; Wu, K.Y. Karl .

In: Statistics and Probability Letters, Vol. 156, 108588, 01.2020, p. 1-6.

Research output: Contribution to journalArticle

TY - JOUR

T1 - A note on discrete multivariate Markov random field models

AU - Ip, H.L.

AU - Wu, K.Y. Karl

PY - 2020/1

Y1 - 2020/1

N2 - Markov random field (MRF) is commonly used in modelling spatially dependent data. These models are often referred to as auto-models. While univariate auto-models have been extensively studied in the literature, discrete multivariate MRF has not attracted much attention. This paper attempts to fill the research gap by providing some results on the discrete multivariate MRF scheme, which forms the theoretical foundation to construct models for spatially dependent categorical data. The results presented in this paper allow the formulation of a novel auto-model and justify the validity of the recently proposed auto-multinomial model.

AB - Markov random field (MRF) is commonly used in modelling spatially dependent data. These models are often referred to as auto-models. While univariate auto-models have been extensively studied in the literature, discrete multivariate MRF has not attracted much attention. This paper attempts to fill the research gap by providing some results on the discrete multivariate MRF scheme, which forms the theoretical foundation to construct models for spatially dependent categorical data. The results presented in this paper allow the formulation of a novel auto-model and justify the validity of the recently proposed auto-multinomial model.

KW - Auto-models

KW - Lattice system

KW - MRF

KW - Spatial correlation

U2 - 10.1016/j.spl.2019.108588

DO - 10.1016/j.spl.2019.108588

M3 - Article

VL - 156

SP - 1

EP - 6

JO - Statistics and Probability Letters

JF - Statistics and Probability Letters

SN - 0167-7152

M1 - 108588

ER -