DNP-XGboostCerete: Satellite-Derived Building Footprints for Cereté Municipality

Estimated building footprint layers generated by applying an XGBoost classification model to satellite imagery for the Municipality of Cereté, Colombia. The model was developed during the DataSandbox Colombia 2020 by the Data Science Unit of the National Planning Department (Departamento Nacional de Planeación). Data is projected in MagnaSirgas 3115 and was last updated on 2026-05-18.

Use Cases

• Mapping urban expansion by visualizing building footprints using the polygon and municipio columns.
• Assessing model confidence for individual building detections by analyzing the probability column.
• Aggregating building counts or densities per administrative area using the municipio and index columns.
• Validating and improving remote sensing classification models by using the polygon geometry and associated probability scores.

Strengths

• Data is derived from a specific machine learning methodology (XGBoost) applied to satellite imagery.
• Includes a confidence metric (probability) for each classified feature.
• Covers a defined geographic area (Municipality of Cereté, Colombia).
• Available in multiple standard geospatial data formats (CSV, JSON, XML, RDF).

Limitations

• Row count and dataset size are unknown, which may limit suitability assessment.
• Description metadata is limited; actual data quality and accuracy require manual inspection after download.
• The license and author information are not specified.

Provenance

Source

Data Science Unit of the National Planning Department (Departamento Nacional de Planeación), Colombia.

Collection Method

XGBoost classification model applied to satellite imagery.

Freshness

Last updated 2026-05-18 16:36:14; freshness should be verified.

Geography

Municipality of Cereté, Colombia.