An application of geographical random forests for population estimation in Dakar, Senegal using very-high-resolution satellite imagery

Stefanos Georganos, Tais Grippa, Assane Gadiaga, Sabine Vanhuysse, Stamatis Kalogirou, Moritz Lennert, Catherine Linard

Résultats de recherche: Contribution dans un livre/un catalogue/un rapport/dans les actes d'une conférenceArticle dans les actes d'une conférence/un colloque

Résumé

In this paper we investigate a local implementation of Random Forest (RF), named Geographical Random Forest (GRF) to predict population density with Very-High-Resolution Remote Sensing (VHHRS) data. As an independent variable we use population density at the neighborhood level from the 2013 census of Dakar, while as explanatory features, the proportions of three different built-up types in each neighborhood derived from a VHHRS land cover classification. The results demonstrated, that by using an appropriate geographic scale to calibrate GRF, we can maximize prediction accuracy due to the incorporation of spatial heterogeneity in the estimates. Additionally, since GRF is an ensemble of local sub-models, the results can be mapped, highlighting local model performance and other interesting spatial variations. Consequently, GRF is suggested as an interesting exploratory and explanatory technique to model remotely-sensed spatially heterogeneous relationships.

langue originaleAnglais
titre2019 Joint Urban Remote Sensing Event (JURSE)
Sous-titreMay, 22-24, 2019, Vannes, France
EditeurInstitute of Electrical and Electronics Engineers Inc.
ISBN (Electronique)9781728100098
Les DOIs
étatPublié - 1 mai 2019
Evénement2019 Joint Urban Remote Sensing Event, JURSE 2019 - Vannes, France
Durée: 22 mai 201924 mai 2019

Série de publications

Nom2019 Joint Urban Remote Sensing Event, JURSE 2019

Une conférence

Une conférence2019 Joint Urban Remote Sensing Event, JURSE 2019
PaysFrance
La villeVannes
période22/05/1924/05/19

Empreinte digitale

Senegal
population estimation
satellite imagery
Satellite imagery
Remote sensing
high resolution
population density
remote sensing
census
land cover
spatial variation
proportion
prediction
estimates
predictions
performance

Citer ceci

Georganos, S., Grippa, T., Gadiaga, A., Vanhuysse, S., Kalogirou, S., Lennert, M., & Linard, C. (2019). An application of geographical random forests for population estimation in Dakar, Senegal using very-high-resolution satellite imagery. Dans 2019 Joint Urban Remote Sensing Event (JURSE): May, 22-24, 2019, Vannes, France [8809049] (2019 Joint Urban Remote Sensing Event, JURSE 2019). Institute of Electrical and Electronics Engineers Inc.. https://doi.org/10.1109/JURSE.2019.8809049
Georganos, Stefanos ; Grippa, Tais ; Gadiaga, Assane ; Vanhuysse, Sabine ; Kalogirou, Stamatis ; Lennert, Moritz ; Linard, Catherine. / An application of geographical random forests for population estimation in Dakar, Senegal using very-high-resolution satellite imagery. 2019 Joint Urban Remote Sensing Event (JURSE): May, 22-24, 2019, Vannes, France. Institute of Electrical and Electronics Engineers Inc., 2019. (2019 Joint Urban Remote Sensing Event, JURSE 2019).
@inproceedings{a461a8d1698f4759bb2c0f75fa15bfc7,
title = "An application of geographical random forests for population estimation in Dakar, Senegal using very-high-resolution satellite imagery",
abstract = "In this paper we investigate a local implementation of Random Forest (RF), named Geographical Random Forest (GRF) to predict population density with Very-High-Resolution Remote Sensing (VHHRS) data. As an independent variable we use population density at the neighborhood level from the 2013 census of Dakar, while as explanatory features, the proportions of three different built-up types in each neighborhood derived from a VHHRS land cover classification. The results demonstrated, that by using an appropriate geographic scale to calibrate GRF, we can maximize prediction accuracy due to the incorporation of spatial heterogeneity in the estimates. Additionally, since GRF is an ensemble of local sub-models, the results can be mapped, highlighting local model performance and other interesting spatial variations. Consequently, GRF is suggested as an interesting exploratory and explanatory technique to model remotely-sensed spatially heterogeneous relationships.",
keywords = "population estimation, random forest, spatial heterogeneity, very-high-resolution remote sensing",
author = "Stefanos Georganos and Tais Grippa and Assane Gadiaga and Sabine Vanhuysse and Stamatis Kalogirou and Moritz Lennert and Catherine Linard",
year = "2019",
month = "5",
day = "1",
doi = "10.1109/JURSE.2019.8809049",
language = "English",
series = "2019 Joint Urban Remote Sensing Event, JURSE 2019",
publisher = "Institute of Electrical and Electronics Engineers Inc.",
booktitle = "2019 Joint Urban Remote Sensing Event (JURSE)",

}

Georganos, S, Grippa, T, Gadiaga, A, Vanhuysse, S, Kalogirou, S, Lennert, M & Linard, C 2019, An application of geographical random forests for population estimation in Dakar, Senegal using very-high-resolution satellite imagery. Dans 2019 Joint Urban Remote Sensing Event (JURSE): May, 22-24, 2019, Vannes, France., 8809049, 2019 Joint Urban Remote Sensing Event, JURSE 2019, Institute of Electrical and Electronics Engineers Inc., 2019 Joint Urban Remote Sensing Event, JURSE 2019, Vannes, France, 22/05/19. https://doi.org/10.1109/JURSE.2019.8809049

An application of geographical random forests for population estimation in Dakar, Senegal using very-high-resolution satellite imagery. / Georganos, Stefanos; Grippa, Tais; Gadiaga, Assane; Vanhuysse, Sabine; Kalogirou, Stamatis; Lennert, Moritz; Linard, Catherine.

2019 Joint Urban Remote Sensing Event (JURSE): May, 22-24, 2019, Vannes, France. Institute of Electrical and Electronics Engineers Inc., 2019. 8809049 (2019 Joint Urban Remote Sensing Event, JURSE 2019).

Résultats de recherche: Contribution dans un livre/un catalogue/un rapport/dans les actes d'une conférenceArticle dans les actes d'une conférence/un colloque

TY - GEN

T1 - An application of geographical random forests for population estimation in Dakar, Senegal using very-high-resolution satellite imagery

AU - Georganos, Stefanos

AU - Grippa, Tais

AU - Gadiaga, Assane

AU - Vanhuysse, Sabine

AU - Kalogirou, Stamatis

AU - Lennert, Moritz

AU - Linard, Catherine

PY - 2019/5/1

Y1 - 2019/5/1

N2 - In this paper we investigate a local implementation of Random Forest (RF), named Geographical Random Forest (GRF) to predict population density with Very-High-Resolution Remote Sensing (VHHRS) data. As an independent variable we use population density at the neighborhood level from the 2013 census of Dakar, while as explanatory features, the proportions of three different built-up types in each neighborhood derived from a VHHRS land cover classification. The results demonstrated, that by using an appropriate geographic scale to calibrate GRF, we can maximize prediction accuracy due to the incorporation of spatial heterogeneity in the estimates. Additionally, since GRF is an ensemble of local sub-models, the results can be mapped, highlighting local model performance and other interesting spatial variations. Consequently, GRF is suggested as an interesting exploratory and explanatory technique to model remotely-sensed spatially heterogeneous relationships.

AB - In this paper we investigate a local implementation of Random Forest (RF), named Geographical Random Forest (GRF) to predict population density with Very-High-Resolution Remote Sensing (VHHRS) data. As an independent variable we use population density at the neighborhood level from the 2013 census of Dakar, while as explanatory features, the proportions of three different built-up types in each neighborhood derived from a VHHRS land cover classification. The results demonstrated, that by using an appropriate geographic scale to calibrate GRF, we can maximize prediction accuracy due to the incorporation of spatial heterogeneity in the estimates. Additionally, since GRF is an ensemble of local sub-models, the results can be mapped, highlighting local model performance and other interesting spatial variations. Consequently, GRF is suggested as an interesting exploratory and explanatory technique to model remotely-sensed spatially heterogeneous relationships.

KW - population estimation

KW - random forest

KW - spatial heterogeneity

KW - very-high-resolution remote sensing

UR - http://www.scopus.com/inward/record.url?scp=85072017863&partnerID=8YFLogxK

U2 - 10.1109/JURSE.2019.8809049

DO - 10.1109/JURSE.2019.8809049

M3 - Conference contribution

AN - SCOPUS:85072017863

T3 - 2019 Joint Urban Remote Sensing Event, JURSE 2019

BT - 2019 Joint Urban Remote Sensing Event (JURSE)

PB - Institute of Electrical and Electronics Engineers Inc.

ER -

Georganos S, Grippa T, Gadiaga A, Vanhuysse S, Kalogirou S, Lennert M et al. An application of geographical random forests for population estimation in Dakar, Senegal using very-high-resolution satellite imagery. Dans 2019 Joint Urban Remote Sensing Event (JURSE): May, 22-24, 2019, Vannes, France. Institute of Electrical and Electronics Engineers Inc. 2019. 8809049. (2019 Joint Urban Remote Sensing Event, JURSE 2019). https://doi.org/10.1109/JURSE.2019.8809049