Learning to cluster urban areas: two competitive approaches and an empirical validation

EPJ Data Science

Regular Article

Gabriel Cuchacovic¹  Axel Reyes²  Jan Dimter³  Hernán Valdivieso³  Marcelo Mendoza⁴  Camila Vera⁴  Hans Löbel⁵  Nicolás Alvarado⁶  Francesca Lucchini⁶  Felipe Gutiérrez⁶  Naim Bro⁷  Sergio Toro⁸ 

[1] Department of Computer Science, Pontificia Universidad Católica de Chile, Santiago, Chile;Department of Computer Science, Pontificia Universidad Católica de Chile, Santiago, Chile;Department of Informatics, Universidad Técnica Federico Santa María, Santiago, Chile;Department of Computer Science, Pontificia Universidad Católica de Chile, Santiago, Chile;Millennium Institute for Foundational Research on Data (IMFD), Santiago, Chile;Department of Computer Science, Pontificia Universidad Católica de Chile, Santiago, Chile;Millennium Institute for Foundational Research on Data (IMFD), Santiago, Chile;National Center of Artificial Intelligence, Santiago, Chile;Department of Computer Science, Pontificia Universidad Católica de Chile, Santiago, Chile;Millennium Institute for Foundational Research on Data (IMFD), Santiago, Chile;National Center of Artificial Intelligence, Santiago, Chile;Department of Transport Engineering and Logistics, Pontificia Universidad Católica de Chile, Santiago, Chile;Department of Computer Science, Pontificia Universidad Católica de Chile, Santiago, Chile;National Center of Artificial Intelligence, Santiago, Chile;Millennium Institute for Foundational Research on Data (IMFD), Santiago, Chile;Faculty of Government, Universidad Adolfo Ibáñez, Santiago, Chile;Millennium Institute for Foundational Research on Data (IMFD), Santiago, Chile;Public Administration and Political Science, Universidad de Concepción, Concepción, Chile;School of Government, Universidad Mayor, Santiago, Chile;

关键词: Urban clustering;    Graph Neural Networks;    Gaussian Mixture Models;   

DOI  :  10.1140/epjds/s13688-022-00374-2

 received in 2022-07-12, accepted in 2022-12-05,  发布年份 2022

来源: Springer

【 摘 要 】

Urban clustering detects geographical units that are internally homogeneous and distinct from their surroundings. It has applications in urban planning, but few studies compare the effectiveness of different methods. We study two techniques that represent two families of urban clustering algorithms: Gaussian Mixture Models (GMMs), which operate on spatially distributed data, and Deep Modularity Networks (DMONs), which work on attributed graphs of proximal nodes. To explore the strengths and limitations of these techniques, we studied their parametric sensitivity under different conditions, considering the spatial resolution, granularity of representation, and the number of descriptive attributes, among other relevant factors. To validate the methods, we asked residents of Santiago, Chile, to respond to a survey comparing city clustering solutions produced using the different methods. Our study shows that DMON is slightly preferred over GMM and that social features seem to be the most important ones to cluster urban areas.

【 授权许可】

CC BY   
© The Author(s) 2022

【 图 表 】

Fig. 10

Fig. 4

Fig. 3

Fig. 2

Fig. 2

Fig. 6

Fig. 5

Fig. 1

Fig. 4

Fig. 4

Fig. 2

Fig. 1

12888_2022_4482_Article_IEq3.gif

12888_2022_4482_Article_IEq2.gif

12888_2022_4482_Article_IEq1.gif

41408_2022_765_Article_IEq1.gif

Fig. 2

12982_2022_119_Article_IEq84.gif

12982_2022_119_Article_IEq83.gif

12982_2022_119_Article_IEq82.gif

12982_2022_119_Article_IEq81.gif

12982_2022_119_Article_IEq80.gif

12982_2022_119_Article_IEq78.gif

12982_2022_119_Article_IEq77.gif

12982_2022_119_Article_IEq76.gif

12982_2022_119_Article_IEq75.gif

12982_2022_119_Article_IEq221.gif

12982_2022_119_Article_IEq219.gif

12982_2022_119_Article_IEq218.gif

12982_2022_119_Article_IEq216.gif

12982_2022_119_Article_IEq214.gif

Fig. 3

Fig. 2

12982_2022_119_Article_IEq210.gif

12982_2022_119_Article_IEq208.gif

12982_2022_119_Article_IEq207.gif

12982_2022_119_Article_IEq28.gif

12982_2022_119_Article_IEq27.gif

12982_2022_119_Article_IEq26.gif

Fig. 1

12982_2022_119_Article_IEq23.gif

Fig. 3

Fig. 2

12888_2022_4322_Article_IEq8.gif

Fig. 4