Combining spatial transition probabilities for stochastic simulation of categorical fields

Guofeng Cao, Phaedon C. Kyriakidis, Michael Goodchild

Research output: Contribution to journalArticle

11 Citations (Scopus)

Abstract

Categorical spatial data, such as land use classes and socioeconomic statistics data, are important data sources in geographical information science (GIS). The investigation of spatial patterns implied in these data can benefit many aspects of GIS research, such as classification of spatial data, spatial data mining, and spatial uncertainty modeling. However, the discrete nature of categorical data limits the application of traditional kriging methods widely used in Gaussian random fields. In this article, we present a new probabilistic method for modeling the posterior probability of class occurrence at any target location in space-given known class labels at source data locations within a neighborhood around that prediction location. In the proposed method, transition probabilities rather than indicator covariances or variograms are used as measures of spatial structure and the conditional or posterior (multi-point) probability is approximated by a weighted combination of preposterior (two-point) transition probabilities, while accounting for spatial interdependencies often ignored by existing approaches. In addition, the connections of the proposed method with probabilistic graphical models (Bayesian networks) and weights of evidence method are also discussed. The advantages of this new proposed approach are analyzed and highlighted through a case study involving the generation of spatial patterns via sequential indicator simulation.

Original languageEnglish (US)
Pages (from-to)1773-1791
Number of pages19
JournalInternational Journal of Geographical Information Science
Volume25
Issue number11
DOIs
StatePublished - Nov 1 2011
Externally publishedYes

Fingerprint

simulation
Information science
spatial data
Bayesian networks
Land use
information science
data mining
Data mining
variogram
Labels
kriging
Statistics
modeling
method
land use
prediction
statistics
uncertainty
indicator
science

Keywords

  • categorical data
  • conditional independence
  • indicator kriging
  • Tau model

ASJC Scopus subject areas

  • Geography, Planning and Development
  • Information Systems
  • Library and Information Sciences

Cite this

Combining spatial transition probabilities for stochastic simulation of categorical fields. / Cao, Guofeng; Kyriakidis, Phaedon C.; Goodchild, Michael.

In: International Journal of Geographical Information Science, Vol. 25, No. 11, 01.11.2011, p. 1773-1791.

Research output: Contribution to journalArticle

@article{fff26221ce024e369e8c59140fd150e2,
title = "Combining spatial transition probabilities for stochastic simulation of categorical fields",
abstract = "Categorical spatial data, such as land use classes and socioeconomic statistics data, are important data sources in geographical information science (GIS). The investigation of spatial patterns implied in these data can benefit many aspects of GIS research, such as classification of spatial data, spatial data mining, and spatial uncertainty modeling. However, the discrete nature of categorical data limits the application of traditional kriging methods widely used in Gaussian random fields. In this article, we present a new probabilistic method for modeling the posterior probability of class occurrence at any target location in space-given known class labels at source data locations within a neighborhood around that prediction location. In the proposed method, transition probabilities rather than indicator covariances or variograms are used as measures of spatial structure and the conditional or posterior (multi-point) probability is approximated by a weighted combination of preposterior (two-point) transition probabilities, while accounting for spatial interdependencies often ignored by existing approaches. In addition, the connections of the proposed method with probabilistic graphical models (Bayesian networks) and weights of evidence method are also discussed. The advantages of this new proposed approach are analyzed and highlighted through a case study involving the generation of spatial patterns via sequential indicator simulation.",
keywords = "categorical data, conditional independence, indicator kriging, Tau model",
author = "Guofeng Cao and Kyriakidis, {Phaedon C.} and Michael Goodchild",
year = "2011",
month = "11",
day = "1",
doi = "10.1080/13658816.2010.528421",
language = "English (US)",
volume = "25",
pages = "1773--1791",
journal = "International Journal of Geographical Information Science",
issn = "1365-8816",
publisher = "Taylor and Francis Ltd.",
number = "11",

}

TY - JOUR

T1 - Combining spatial transition probabilities for stochastic simulation of categorical fields

AU - Cao, Guofeng

AU - Kyriakidis, Phaedon C.

AU - Goodchild, Michael

PY - 2011/11/1

Y1 - 2011/11/1

N2 - Categorical spatial data, such as land use classes and socioeconomic statistics data, are important data sources in geographical information science (GIS). The investigation of spatial patterns implied in these data can benefit many aspects of GIS research, such as classification of spatial data, spatial data mining, and spatial uncertainty modeling. However, the discrete nature of categorical data limits the application of traditional kriging methods widely used in Gaussian random fields. In this article, we present a new probabilistic method for modeling the posterior probability of class occurrence at any target location in space-given known class labels at source data locations within a neighborhood around that prediction location. In the proposed method, transition probabilities rather than indicator covariances or variograms are used as measures of spatial structure and the conditional or posterior (multi-point) probability is approximated by a weighted combination of preposterior (two-point) transition probabilities, while accounting for spatial interdependencies often ignored by existing approaches. In addition, the connections of the proposed method with probabilistic graphical models (Bayesian networks) and weights of evidence method are also discussed. The advantages of this new proposed approach are analyzed and highlighted through a case study involving the generation of spatial patterns via sequential indicator simulation.

AB - Categorical spatial data, such as land use classes and socioeconomic statistics data, are important data sources in geographical information science (GIS). The investigation of spatial patterns implied in these data can benefit many aspects of GIS research, such as classification of spatial data, spatial data mining, and spatial uncertainty modeling. However, the discrete nature of categorical data limits the application of traditional kriging methods widely used in Gaussian random fields. In this article, we present a new probabilistic method for modeling the posterior probability of class occurrence at any target location in space-given known class labels at source data locations within a neighborhood around that prediction location. In the proposed method, transition probabilities rather than indicator covariances or variograms are used as measures of spatial structure and the conditional or posterior (multi-point) probability is approximated by a weighted combination of preposterior (two-point) transition probabilities, while accounting for spatial interdependencies often ignored by existing approaches. In addition, the connections of the proposed method with probabilistic graphical models (Bayesian networks) and weights of evidence method are also discussed. The advantages of this new proposed approach are analyzed and highlighted through a case study involving the generation of spatial patterns via sequential indicator simulation.

KW - categorical data

KW - conditional independence

KW - indicator kriging

KW - Tau model

UR - http://www.scopus.com/inward/record.url?scp=84859122653&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=84859122653&partnerID=8YFLogxK

U2 - 10.1080/13658816.2010.528421

DO - 10.1080/13658816.2010.528421

M3 - Article

VL - 25

SP - 1773

EP - 1791

JO - International Journal of Geographical Information Science

JF - International Journal of Geographical Information Science

SN - 1365-8816

IS - 11

ER -