Monitoring forest health using hyperspectral imagery: Does feature selection improve the performance of machine-learning techniques?

This is a research compendium (RC) for the publication

Monitoring forest health using hyperspectral imagery: Does feature selection improve the performance of machine-learning techniques?

Code, figures, appendices and the manuscript can be found in the corresponding GitHub repository.

This RC is a static snapshot at the time of submission. The GitHub repository holds the latest version and may see changes after the publication was accepted.

Data sources and description

• aoi.gpkg: Area of interest for downloading Sentinel-2 images. Not used in the publication. Source: Custom.
• forest_mask.gpkg: A forest/non-forest mask of the Basque Country. Not used in the publication. Source: Custom.
• hyperspectral.zip: Hyperspectral remote sensing data used to extract reflectance values on the tree level. Source: Custom.
• plot-locations.gpkg: Spatial location of the plots used in the study. Source: Custom.
• tree-in-situ-data-corrected.zip: Corrected in-situ data containing defoliation information on the tree level. A correction of the spatial location was applied by the creators of the data. Source: Custom.
• tree-in-situ-data.zip: First version of in-situ data containing defoliation information on the tree level. Not used in the publication. Source: Custom.

Licenses

All files are licensed under CC BY 4.0.