Search Results

Documents authored by Harris, Paul

Document
Short Paper
DOI: 10.4230/LIPIcs.GIScience.2023.17
Smarter Than Your Average Model - Bayesian Model Averaging as a Spatial Analysis Tool (Short Paper)

Authors: Chris Brunsdon, Paul Harris, and Alexis Comber

Published in: LIPIcs, Volume 277, 12th International Conference on Geographic Information Science (GIScience 2023)

Abstract
Bayesian modelling averaging (BMA) allows the results of analysing competing data models to be combined, and the relative plausibility of the models to be assessed. Here, the potential to apply this approach to spatial statistical models is considered, using an example of spatially varying coefficient modelling applied to data from the 2016 UK referendum on leaving the EU.
Cite as

Chris Brunsdon, Paul Harris, and Alexis Comber. Smarter Than Your Average Model - Bayesian Model Averaging as a Spatial Analysis Tool (Short Paper). In 12th International Conference on Geographic Information Science (GIScience 2023). Leibniz International Proceedings in Informatics (LIPIcs), Volume 277, pp. 17:1-17:6, Schloss Dagstuhl – Leibniz-Zentrum für Informatik (2023)

Copy BibTex To Clipboard

@InProceedings{brunsdon_et_al:LIPIcs.GIScience.2023.17,
  author =	{Brunsdon, Chris and Harris, Paul and Comber, Alexis},
  title =	{{Smarter Than Your Average Model - Bayesian Model Averaging as a Spatial Analysis Tool}},
  booktitle =	{12th International Conference on Geographic Information Science (GIScience 2023)},
  pages =	{17:1--17:6},
  series =	{Leibniz International Proceedings in Informatics (LIPIcs)},
  ISBN =	{978-3-95977-288-4},
  ISSN =	{1868-8969},
  year =	{2023},
  volume =	{277},
  editor =	{Beecham, Roger and Long, Jed A. and Smith, Dianna and Zhao, Qunshan and Wise, Sarah},
  publisher =	{Schloss Dagstuhl -- Leibniz-Zentrum f{\"u}r Informatik},
  address =	{Dagstuhl, Germany},
  URL =		{https://drops-dev.dagstuhl.de/entities/document/10.4230/LIPIcs.GIScience.2023.17},
  URN =		{urn:nbn:de:0030-drops-189123},
  doi =		{10.4230/LIPIcs.GIScience.2023.17},
  annote =	{Keywords: Bayesian, Varying coefficient regression, Spatial statistics}
}
Document
Short Paper
DOI: 10.4230/LIPIcs.GIScience.2023.22
Multiscale Spatially and Temporally Varying Coefficient Modelling Using a Geographic and Temporal Gaussian Process GAM (GTGP-GAM) (Short Paper)

Authors: Alexis Comber, Paul Harris, and Chris Brunsdon

Published in: LIPIcs, Volume 277, 12th International Conference on Geographic Information Science (GIScience 2023)

Abstract
The paper develops a novel approach to spatially and temporally varying coefficient (STVC) modelling, using Generalised Additive Models (GAMs) with Gaussian Process (GP) splines parameterised with location and time variables - a Geographic and Temporal Gaussian Process GAM (GTGP-GAM). This was applied to a Mongolian livestock case study and different forms of GTGP splines were evaluated in which space and time were combined or treated separately. A single 3-D spline with rescaled temporal and spatial attributes resulted in the best model under an assumption that for spatial and temporal processes interact a case studies with a sufficiently large spatial extent is needed. A fully tuned model was then created and the spline smoothing parameters were shown to indicate the degree of variation in covariate spatio-temporal interactions with the target variable.
Cite as

Alexis Comber, Paul Harris, and Chris Brunsdon. Multiscale Spatially and Temporally Varying Coefficient Modelling Using a Geographic and Temporal Gaussian Process GAM (GTGP-GAM) (Short Paper). In 12th International Conference on Geographic Information Science (GIScience 2023). Leibniz International Proceedings in Informatics (LIPIcs), Volume 277, pp. 22:1-22:6, Schloss Dagstuhl – Leibniz-Zentrum für Informatik (2023)

Copy BibTex To Clipboard

@InProceedings{comber_et_al:LIPIcs.GIScience.2023.22,
  author =	{Comber, Alexis and Harris, Paul and Brunsdon, Chris},
  title =	{{Multiscale Spatially and Temporally Varying Coefficient Modelling Using a Geographic and Temporal Gaussian Process GAM (GTGP-GAM)}},
  booktitle =	{12th International Conference on Geographic Information Science (GIScience 2023)},
  pages =	{22:1--22:6},
  series =	{Leibniz International Proceedings in Informatics (LIPIcs)},
  ISBN =	{978-3-95977-288-4},
  ISSN =	{1868-8969},
  year =	{2023},
  volume =	{277},
  editor =	{Beecham, Roger and Long, Jed A. and Smith, Dianna and Zhao, Qunshan and Wise, Sarah},
  publisher =	{Schloss Dagstuhl -- Leibniz-Zentrum f{\"u}r Informatik},
  address =	{Dagstuhl, Germany},
  URL =		{https://drops-dev.dagstuhl.de/entities/document/10.4230/LIPIcs.GIScience.2023.22},
  URN =		{urn:nbn:de:0030-drops-189173},
  doi =		{10.4230/LIPIcs.GIScience.2023.22},
  annote =	{Keywords: Spatial Analysis, Spatiotemproal Analysis}
}
Document
Short Paper
DOI: 10.4230/LIPIcs.COSIT.2022.13
Geographically Varying Coefficient Regression: GWR-Exit and GAM-On? (Short Paper)

Authors: Alexis Comber, Paul Harris, Daisuke Murakami, Narumasa Tsutsumida, and Chris Brunsdon

Published in: LIPIcs, Volume 240, 15th International Conference on Spatial Information Theory (COSIT 2022)

Abstract
This paper describes initial work exploring two spatially varying coefficient models: multi-scale GWR and GAM Gaussian Process spline parameterised by observation location. Both approaches accommodate process spatial heterogeneity and both generate outputs that can be mapped indicating the nature of the process heterogeneity. However the nature of the process heterogeneity they each describe are very different. This suggests that the underlying semantics of such models need to be considered in order to refine the specificity of the questions that are asked of data: simply seeking to understand process spatial heterogeneity may be too semantically coarse.
Cite as

Alexis Comber, Paul Harris, Daisuke Murakami, Narumasa Tsutsumida, and Chris Brunsdon. Geographically Varying Coefficient Regression: GWR-Exit and GAM-On? (Short Paper). In 15th International Conference on Spatial Information Theory (COSIT 2022). Leibniz International Proceedings in Informatics (LIPIcs), Volume 240, pp. 13:1-13:10, Schloss Dagstuhl – Leibniz-Zentrum für Informatik (2022)

Copy BibTex To Clipboard

@InProceedings{comber_et_al:LIPIcs.COSIT.2022.13,
  author =	{Comber, Alexis and Harris, Paul and Murakami, Daisuke and Tsutsumida, Narumasa and Brunsdon, Chris},
  title =	{{Geographically Varying Coefficient Regression: GWR-Exit and GAM-On?}},
  booktitle =	{15th International Conference on Spatial Information Theory (COSIT 2022)},
  pages =	{13:1--13:10},
  series =	{Leibniz International Proceedings in Informatics (LIPIcs)},
  ISBN =	{978-3-95977-257-0},
  ISSN =	{1868-8969},
  year =	{2022},
  volume =	{240},
  editor =	{Ishikawa, Toru and Fabrikant, Sara Irina and Winter, Stephan},
  publisher =	{Schloss Dagstuhl -- Leibniz-Zentrum f{\"u}r Informatik},
  address =	{Dagstuhl, Germany},
  URL =		{https://drops-dev.dagstuhl.de/entities/document/10.4230/LIPIcs.COSIT.2022.13},
  URN =		{urn:nbn:de:0030-drops-168986},
  doi =		{10.4230/LIPIcs.COSIT.2022.13},
  annote =	{Keywords: Geographically weighted regression, Spatial Analysis, Process Spatial Heterogeneity, Model Semantics}
}
Document
Short Paper
DOI: 10.4230/LIPIcs.COSIT.2022.21
A Comparison of Geographically Weighted Principal Components Analysis Methodologies (Short Paper)

Authors: Narumasa Tsutsumida, Daisuke Murakami, Takahiro Yoshida, Tomoki Nakaya, Binbin Lu, Paul Harris, and Alexis Comber

Published in: LIPIcs, Volume 240, 15th International Conference on Spatial Information Theory (COSIT 2022)

Abstract
Principal components analysis (PCA) is a useful analytical tool to represent key characteristics of multivariate data, but does not account for spatial effects when applied in geographical situations. A geographically weighted PCA (GWPCA) caters to this issue, specifically in terms of capturing spatial heterogeneity. However, in certain situations, a GWPCA provides outputs that vary discontinuously spatially, which are difficult to interpret and are not associated with the output from a conventional (global) PCA any more. This study underlines a GW non-negative PCA, a geographically weighted version of non-negative PCA, to overcome this issue by constraining loading values non-negatively. Case study results with a complex multivariate spatial dataset demonstrate such benefits, where GW non-negative PCA allows improved interpretations than that found with conventional GWPCA.
Cite as

Narumasa Tsutsumida, Daisuke Murakami, Takahiro Yoshida, Tomoki Nakaya, Binbin Lu, Paul Harris, and Alexis Comber. A Comparison of Geographically Weighted Principal Components Analysis Methodologies (Short Paper). In 15th International Conference on Spatial Information Theory (COSIT 2022). Leibniz International Proceedings in Informatics (LIPIcs), Volume 240, pp. 21:1-21:6, Schloss Dagstuhl – Leibniz-Zentrum für Informatik (2022)

Copy BibTex To Clipboard

@InProceedings{tsutsumida_et_al:LIPIcs.COSIT.2022.21,
  author =	{Tsutsumida, Narumasa and Murakami, Daisuke and Yoshida, Takahiro and Nakaya, Tomoki and Lu, Binbin and Harris, Paul and Comber, Alexis},
  title =	{{A Comparison of Geographically Weighted Principal Components Analysis Methodologies}},
  booktitle =	{15th International Conference on Spatial Information Theory (COSIT 2022)},
  pages =	{21:1--21:6},
  series =	{Leibniz International Proceedings in Informatics (LIPIcs)},
  ISBN =	{978-3-95977-257-0},
  ISSN =	{1868-8969},
  year =	{2022},
  volume =	{240},
  editor =	{Ishikawa, Toru and Fabrikant, Sara Irina and Winter, Stephan},
  publisher =	{Schloss Dagstuhl -- Leibniz-Zentrum f{\"u}r Informatik},
  address =	{Dagstuhl, Germany},
  URL =		{https://drops-dev.dagstuhl.de/entities/document/10.4230/LIPIcs.COSIT.2022.21},
  URN =		{urn:nbn:de:0030-drops-169062},
  doi =		{10.4230/LIPIcs.COSIT.2022.21},
  annote =	{Keywords: Spatial heterogeneity, Geographically weighted, Sparsity, PCA}
}
Questions / Remarks / Feedback
X

Feedback for Dagstuhl Publishing


Thanks for your feedback!

Feedback submitted

Could not send message

Please try again later or send an E-mail
© 2023-2024 Schloss Dagstuhl – LZI GmbH Imprint Privacy Contact