04 · Digital soil mapping with machine learning

Digital Soil Mapping
Regression-matrix, covariate selection and models (random forest, quantile regression forest, Cubist).

Duration: 4 h · Level: Intermediate

Open in the SoilFER Lab

WarningContent being imported

This module will be populated from the SoilFER Capacity Development training material (FAO-SID/SoilFER/Capacity Development). See scripts/import-materials.sh.

Learning objectives

By the end of this module you will be able to:

  • build the regression matrix from points and covariates;
  • select covariates and tune machine-learning models;
  • predict soil properties over the whole country.
Back to top