The project aims to develop a classification and regression module to analyze agricultural data collected during experimental phases conducted either in the laboratory or in the field.

The goal is to build predictive models capable of estimating the mineralization of soil organic matter and understanding how environmental factors influence these outcomes. A regression tree-based model was adopted, trained on real experimental data, to iteratively partition variables into homogeneous intervals relative to the target variable. 

This approach enables the identification of key factors affecting mineralization and allows precise prediction of expected values for soil organic matter. The method provides an interpretable and practical solution that offers reliable predictions based on easily accessible data, thereby supporting targeted decision-making in agricultural contexts.