Atomic Spatial Processes

Sean Jewell, Neil Spencer, Alexandre Bouchard-Côté
Proceedings of the 32nd International Conference on Machine Learning, PMLR 37:248-256, 2015.

Abstract

The emergence of compact GPS systems and the establishment of open data initiatives has resulted in widespread availability of spatial data for many urban centres. These data can be leveraged to develop data-driven intelligent resource allocation systems for urban issues such as policing, sanitation, and transportation. We employ techniques from Bayesian non-parametric statistics to develop a process which captures a common characteristic of urban spatial datasets. Specifically, our new spatial process framework models events which occur repeatedly at discrete spatial points, the number and locations of which are unknown a priori. We develop a representation of our spatial process which facilitates posterior simulation, resulting in an interpretable and computationally tractable model. The framework’s superiority over both empirical grid-based models and Dirichlet process mixture models is demonstrated by fitting, interpreting, and comparing models of graffiti prevalence for both downtown Vancouver and Manhattan.

Cite this Paper

BibTeX
@InProceedings{pmlr-v37-jewell15, title = {Atomic Spatial Processes}, author = {Jewell, Sean and Spencer, Neil and Bouchard-Côté, Alexandre}, booktitle = {Proceedings of the 32nd International Conference on Machine Learning}, pages = {248--256}, year = {2015}, editor = {Bach, Francis and Blei, David}, volume = {37}, series = {Proceedings of Machine Learning Research}, address = {Lille, France}, month = {07--09 Jul}, publisher = {PMLR}, pdf = {http://proceedings.mlr.press/v37/jewell15.pdf}, url = {https://proceedings.mlr.press/v37/jewell15.html}, abstract = {The emergence of compact GPS systems and the establishment of open data initiatives has resulted in widespread availability of spatial data for many urban centres. These data can be leveraged to develop data-driven intelligent resource allocation systems for urban issues such as policing, sanitation, and transportation. We employ techniques from Bayesian non-parametric statistics to develop a process which captures a common characteristic of urban spatial datasets. Specifically, our new spatial process framework models events which occur repeatedly at discrete spatial points, the number and locations of which are unknown a priori. We develop a representation of our spatial process which facilitates posterior simulation, resulting in an interpretable and computationally tractable model. The framework’s superiority over both empirical grid-based models and Dirichlet process mixture models is demonstrated by fitting, interpreting, and comparing models of graffiti prevalence for both downtown Vancouver and Manhattan.} }
Endnote
%0 Conference Paper %T Atomic Spatial Processes %A Sean Jewell %A Neil Spencer %A Alexandre Bouchard-Côté %B Proceedings of the 32nd International Conference on Machine Learning %C Proceedings of Machine Learning Research %D 2015 %E Francis Bach %E David Blei %F pmlr-v37-jewell15 %I PMLR %P 248--256 %U https://proceedings.mlr.press/v37/jewell15.html %V 37 %X The emergence of compact GPS systems and the establishment of open data initiatives has resulted in widespread availability of spatial data for many urban centres. These data can be leveraged to develop data-driven intelligent resource allocation systems for urban issues such as policing, sanitation, and transportation. We employ techniques from Bayesian non-parametric statistics to develop a process which captures a common characteristic of urban spatial datasets. Specifically, our new spatial process framework models events which occur repeatedly at discrete spatial points, the number and locations of which are unknown a priori. We develop a representation of our spatial process which facilitates posterior simulation, resulting in an interpretable and computationally tractable model. The framework’s superiority over both empirical grid-based models and Dirichlet process mixture models is demonstrated by fitting, interpreting, and comparing models of graffiti prevalence for both downtown Vancouver and Manhattan.
RIS
TY - CPAPER TI - Atomic Spatial Processes AU - Sean Jewell AU - Neil Spencer AU - Alexandre Bouchard-Côté BT - Proceedings of the 32nd International Conference on Machine Learning DA - 2015/06/01 ED - Francis Bach ED - David Blei ID - pmlr-v37-jewell15 PB - PMLR DP - Proceedings of Machine Learning Research VL - 37 SP - 248 EP - 256 L1 - http://proceedings.mlr.press/v37/jewell15.pdf UR - https://proceedings.mlr.press/v37/jewell15.html AB - The emergence of compact GPS systems and the establishment of open data initiatives has resulted in widespread availability of spatial data for many urban centres. These data can be leveraged to develop data-driven intelligent resource allocation systems for urban issues such as policing, sanitation, and transportation. We employ techniques from Bayesian non-parametric statistics to develop a process which captures a common characteristic of urban spatial datasets. Specifically, our new spatial process framework models events which occur repeatedly at discrete spatial points, the number and locations of which are unknown a priori. We develop a representation of our spatial process which facilitates posterior simulation, resulting in an interpretable and computationally tractable model. The framework’s superiority over both empirical grid-based models and Dirichlet process mixture models is demonstrated by fitting, interpreting, and comparing models of graffiti prevalence for both downtown Vancouver and Manhattan. ER -
APA
Jewell, S., Spencer, N. & Bouchard-Côté, A.. (2015). Atomic Spatial Processes. Proceedings of the 32nd International Conference on Machine Learning, in Proceedings of Machine Learning Research 37:248-256 Available from https://proceedings.mlr.press/v37/jewell15.html.

Related Material