Prediction of Soil Salinity/Sodicity and Salt-Affected Soil Classes from Soluble Salt Ions Using Machine Learning Algorithms

Salt-affected soils are related to salinity (high content of soluble salts) and/or sodicity (excess sodium), which major leading causes agricultural land degradation. This study aimed evaluate the performances three machine learning (ML) algorithms in predicting soil exchangeable sodium percentage (ESP), electrical conductivity (ECe), and salt-affected classes, from salt ions. The assessed ML models were Partial Least-Squares (PLS), Support Vector Machines (SVM), Random Forests (RF). Soil samples collected High Valley Cochabamba (Bolivia). explanatory variables ions (Na+, K+, Ca2+, Mg2+, HCO3−, Cl−, CO32−, SO42−). be explained comprised ECe ESP, a categorical variable classified through US Salinity Lab criteria. According model validation, SVM RF regressions performed best for estimating ECe, as well ESP. algorithm was superior categories. Soluble Na+ most relevant all predictions, followed by HCO3−. can used predict Additional features might improve models’ performance. obtained may contribute monitoring management area.