Predicting the soil organic carbon by recent machine learning algorithms

In this paper, we focus on the Soil Organic Carbon (SOC) prediction, to discuss a comparative analysis between two recently proposed techniques, namely the Adaptive Network-based Fuzzy Inference System (ANFIS) with fractional Tikhonov regularization and the Extreme Learning Machine (ELM) with the same kind of regularization. Three groups of experiments were performed using some publicly available datasets, in particular from Estonia. The results showed the good accuracy of the ANFIS-based approach.