Prediction and simulation in categorical fields

A transition probability combination approach

Guofeng Cao, Phaedon Kyriakidis, Michael Goodchild

Research output: Chapter in Book/Report/Conference proceedingConference contribution

4 Citations (Scopus)

Abstract

The investigation of spatial patterns implied in categorical spatial data, such as land use and land cover (LULC) classes and socio-economic statistics data, is involved in many aspects of geographical information science, such as spatial uncertainty modeling and spatial data mining. The discrete nature of categorical fields limits the application of traditional analytical methods, such as kriging-type algorithms, widely used in Gaussian random fields. This paper presents a new probabilistic method for modeling the posterior probabilities of class occurrence at any location in space given known class labels at data locations within a neighborhood around that prediction location. In the proposed method, the conditional or posterior (multi-point) probabilities are approximated by weighted combinations of pre-posterior (two-point) transition probabilities (rather than indicator covariances or vari-ograms) while accounting for spatial interdependencies that most of current approaches often ignore. Using sequential indicator simulation based on the properties of a truncated multi-variate Gaussian field as reference, the advantages and disadvantages of this new proposed approach are analyzed and highlighted.

Original languageEnglish (US)
Title of host publication17th ACM SIGSPATIAL International Conference on Advances in Geographic Information Systems, ACM SIGSPATIAL GIS 2009
Pages496-499
Number of pages4
DOIs
StatePublished - Dec 1 2009
Externally publishedYes
Event17th ACM SIGSPATIAL International Conference on Advances in Geographic Information Systems, ACM SIGSPATIAL GIS 2009 - Seattle, WA, United States
Duration: Nov 4 2009Nov 6 2009

Other

Other17th ACM SIGSPATIAL International Conference on Advances in Geographic Information Systems, ACM SIGSPATIAL GIS 2009
CountryUnited States
CitySeattle, WA
Period11/4/0911/6/09

Fingerprint

Transition Probability
Categorical
spatial data
Prediction
prediction
Spatial Data Mining
simulation
Spatial Modeling
Uncertainty Modeling
Gaussian Fields
Variogram
Gaussian Random Field
Simulation
Information science
Land Cover
Nominal or categorical data
Interdependencies
data mining
Probabilistic Methods
Kriging

Keywords

  • Categorical data
  • Conditional independence
  • Tau model

ASJC Scopus subject areas

  • Computer Graphics and Computer-Aided Design
  • Computer Science Applications
  • Information Systems
  • Earth-Surface Processes
  • Modeling and Simulation

Cite this

Cao, G., Kyriakidis, P., & Goodchild, M. (2009). Prediction and simulation in categorical fields: A transition probability combination approach. In 17th ACM SIGSPATIAL International Conference on Advances in Geographic Information Systems, ACM SIGSPATIAL GIS 2009 (pp. 496-499) https://doi.org/10.1145/1653771.1653853

Prediction and simulation in categorical fields : A transition probability combination approach. / Cao, Guofeng; Kyriakidis, Phaedon; Goodchild, Michael.

17th ACM SIGSPATIAL International Conference on Advances in Geographic Information Systems, ACM SIGSPATIAL GIS 2009. 2009. p. 496-499.

Research output: Chapter in Book/Report/Conference proceedingConference contribution

Cao, G, Kyriakidis, P & Goodchild, M 2009, Prediction and simulation in categorical fields: A transition probability combination approach. in 17th ACM SIGSPATIAL International Conference on Advances in Geographic Information Systems, ACM SIGSPATIAL GIS 2009. pp. 496-499, 17th ACM SIGSPATIAL International Conference on Advances in Geographic Information Systems, ACM SIGSPATIAL GIS 2009, Seattle, WA, United States, 11/4/09. https://doi.org/10.1145/1653771.1653853
Cao G, Kyriakidis P, Goodchild M. Prediction and simulation in categorical fields: A transition probability combination approach. In 17th ACM SIGSPATIAL International Conference on Advances in Geographic Information Systems, ACM SIGSPATIAL GIS 2009. 2009. p. 496-499 https://doi.org/10.1145/1653771.1653853
Cao, Guofeng ; Kyriakidis, Phaedon ; Goodchild, Michael. / Prediction and simulation in categorical fields : A transition probability combination approach. 17th ACM SIGSPATIAL International Conference on Advances in Geographic Information Systems, ACM SIGSPATIAL GIS 2009. 2009. pp. 496-499
@inproceedings{e03aa494ca6148bc80cef985697b48d0,
title = "Prediction and simulation in categorical fields: A transition probability combination approach",
abstract = "The investigation of spatial patterns implied in categorical spatial data, such as land use and land cover (LULC) classes and socio-economic statistics data, is involved in many aspects of geographical information science, such as spatial uncertainty modeling and spatial data mining. The discrete nature of categorical fields limits the application of traditional analytical methods, such as kriging-type algorithms, widely used in Gaussian random fields. This paper presents a new probabilistic method for modeling the posterior probabilities of class occurrence at any location in space given known class labels at data locations within a neighborhood around that prediction location. In the proposed method, the conditional or posterior (multi-point) probabilities are approximated by weighted combinations of pre-posterior (two-point) transition probabilities (rather than indicator covariances or vari-ograms) while accounting for spatial interdependencies that most of current approaches often ignore. Using sequential indicator simulation based on the properties of a truncated multi-variate Gaussian field as reference, the advantages and disadvantages of this new proposed approach are analyzed and highlighted.",
keywords = "Categorical data, Conditional independence, Tau model",
author = "Guofeng Cao and Phaedon Kyriakidis and Michael Goodchild",
year = "2009",
month = "12",
day = "1",
doi = "10.1145/1653771.1653853",
language = "English (US)",
isbn = "9781605586496",
pages = "496--499",
booktitle = "17th ACM SIGSPATIAL International Conference on Advances in Geographic Information Systems, ACM SIGSPATIAL GIS 2009",

}

TY - GEN

T1 - Prediction and simulation in categorical fields

T2 - A transition probability combination approach

AU - Cao, Guofeng

AU - Kyriakidis, Phaedon

AU - Goodchild, Michael

PY - 2009/12/1

Y1 - 2009/12/1

N2 - The investigation of spatial patterns implied in categorical spatial data, such as land use and land cover (LULC) classes and socio-economic statistics data, is involved in many aspects of geographical information science, such as spatial uncertainty modeling and spatial data mining. The discrete nature of categorical fields limits the application of traditional analytical methods, such as kriging-type algorithms, widely used in Gaussian random fields. This paper presents a new probabilistic method for modeling the posterior probabilities of class occurrence at any location in space given known class labels at data locations within a neighborhood around that prediction location. In the proposed method, the conditional or posterior (multi-point) probabilities are approximated by weighted combinations of pre-posterior (two-point) transition probabilities (rather than indicator covariances or vari-ograms) while accounting for spatial interdependencies that most of current approaches often ignore. Using sequential indicator simulation based on the properties of a truncated multi-variate Gaussian field as reference, the advantages and disadvantages of this new proposed approach are analyzed and highlighted.

AB - The investigation of spatial patterns implied in categorical spatial data, such as land use and land cover (LULC) classes and socio-economic statistics data, is involved in many aspects of geographical information science, such as spatial uncertainty modeling and spatial data mining. The discrete nature of categorical fields limits the application of traditional analytical methods, such as kriging-type algorithms, widely used in Gaussian random fields. This paper presents a new probabilistic method for modeling the posterior probabilities of class occurrence at any location in space given known class labels at data locations within a neighborhood around that prediction location. In the proposed method, the conditional or posterior (multi-point) probabilities are approximated by weighted combinations of pre-posterior (two-point) transition probabilities (rather than indicator covariances or vari-ograms) while accounting for spatial interdependencies that most of current approaches often ignore. Using sequential indicator simulation based on the properties of a truncated multi-variate Gaussian field as reference, the advantages and disadvantages of this new proposed approach are analyzed and highlighted.

KW - Categorical data

KW - Conditional independence

KW - Tau model

UR - http://www.scopus.com/inward/record.url?scp=74049106930&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=74049106930&partnerID=8YFLogxK

U2 - 10.1145/1653771.1653853

DO - 10.1145/1653771.1653853

M3 - Conference contribution

SN - 9781605586496

SP - 496

EP - 499

BT - 17th ACM SIGSPATIAL International Conference on Advances in Geographic Information Systems, ACM SIGSPATIAL GIS 2009

ER -