Improving digital soil mapping for imbalanced classes through feature selection and resampling
- 1Department of Soil Science, College of Agriculture, Shiraz University
- 2Department of Soil Science, College of Agriculture and Natural Resources, Shahid Bahonar University of Kerman
Abstract
Accurate classification of imbalanced soil data remains a key challenge in Digital Soil Mapping (DSM). We propose a two-step preprocessing strategy that integrates feature selection with resampling to enhance soil class prediction. Using 140 soil profiles from the Darab region (Iran), four feature selection methods (VIF, PCA, Boruta, Expert Viewpoint) were combined with the Synthetic Minority Oversampling Technique (SMOTE) and evaluated with three classifiers (C5.0, Random Forest, XGBoost). Among the tested combinations, VIF + SMOTE with the Random Forest (RF) model achieved the best performance, increasing macro F1 by 15.14% and raising the number of detectable classes from four to five. These gains were class-dependent: SMOTE improved detection for classes with stable environmental signatures (e.g., Haplosalids) but failed for geopedologically complex classes (e.g., Calciusterts). This suggests that post-resampling sample size alone is insufficient and that covariate separability and environmental context are decisive factors. Balancing also shifted variable importance toward topographic and geomorphological predictors, while reducing reliance on weaker covariates. To better address hard-to-separate classes, we recommend boundary-aware resampling (e.g., Borderline-SMOTE, SMOTE-ENN, ADASYN) and covariate enrichment (e.g., horizon descriptors). The resulting soil maps revealed clearer spatial differentiation of rare but management-relevant classes, thereby supporting more targeted and sustainable land-use decisions.
Related
- Mapping soil particle fractions by training digital soil mapping models with surrogate measurements obtained from laboratory and satellite Vis-NIR spectral data
- Machine learning surrogates for flash flood hazard mapping: CSI-optimized feature selection and spatial transferability
- CaSSIS Resampled Data Record (CaRDR): A Near‐Global Multispectral View of Mars From CRISM‐CaSSIS Resampling
- Classification of imbalanced municipal solid waste image datasets
- Key soil attributes and land-management dataset for Australian agroecosystems linking baseline and resampling surveys
BibTeX
@article{sadeghizadeh2026improving,
title = {{Improving digital soil mapping for imbalanced classes through feature selection and resampling}},
author = {Vahideh Sadeghizadeh and Seyed Ali Abtahi and Majid Baghernejad and Azam Jafari and Ali Akbar Moosavi},
journal = {Scientific Reports},
year = {2026},
month = sep,
doi = {10.1038/s41598-026-65549-7},
url = {https://doi.org/10.1038/s41598-026-65549-7},
note = {CC BY},
}
RIS
TY - JOUR
TI - Improving digital soil mapping for imbalanced classes through feature selection and resampling
AU - Vahideh Sadeghizadeh
AU - Seyed Ali Abtahi
AU - Majid Baghernejad
AU - Azam Jafari
AU - Ali Akbar Moosavi
PY - 2026
DA - 2026/09/01
JO - Scientific Reports
DO - 10.1038/s41598-026-65549-7
UR - https://doi.org/10.1038/s41598-026-65549-7
AB - Accurate classification of imbalanced soil data remains a key challenge in Digital Soil Mapping (DSM). We propose a two-step preprocessing strategy that integrates feature selection with resampling to enhance soil class prediction. Using 140 soil profiles from the Darab region (Iran), four feature selection methods (VIF, PCA, Boruta, Expert Viewpoint) were combined with the Synthetic Minority Oversampling Technique (SMOTE) and evaluated with three classifiers (C5.0, Random Forest, XGBoost). Among the tested combinations, VIF + SMOTE with the Random Forest (RF) model achieved the best performance, increasing macro F1 by 15.14% and raising the number of detectable classes from four to five. These gains were class-dependent: SMOTE improved detection for classes with stable environmental signatures (e.g., Haplosalids) but failed for geopedologically complex classes (e.g., Calciusterts). This suggests that post-resampling sample size alone is insufficient and that covariate separability and environmental context are decisive factors. Balancing also shifted variable importance toward topographic and geomorphological predictors, while reducing reliance on weaker covariates. To better address hard-to-separate classes, we recommend boundary-aware resampling (e.g., Borderline-SMOTE, SMOTE-ENN, ADASYN) and covariate enrichment (e.g., horizon descriptors). The resulting soil maps revealed clearer spatial differentiation of rare but management-relevant classes, thereby supporting more targeted and sustainable land-use decisions.
SN - 2045-2322
ER -