Using Near-Infrared Spectroscopy and Stacked Regression for the Simultaneous Determination of Fresh Cattle and Poultry Manure Chemical Properties

Excessive use of animal manure as fertilizers can lead to pollution through the introduction nitrogen, phosphorus, and other mineral compounds environment. Wet chemical analytical methods are traditionally used determine precise composition manage application waste soil. However, such require significant resources carry out processes. Affordable, rapid, accurate analyses various components present in manure, therefore, valuable managing soil mitigating water pollution. In this study, a stacked regression ensemble approach using near-infrared spectroscopy was developed simultaneously amount dry matter, total ammonium phosphorus pentoxide, calcium oxide, magnesium potassium oxide contents both cattle poultry collected from livestock production areas France Reunion Island. The performance regression, an learning algorithm that is formed by collating well-performing models for prediction then compared with machine techniques, including support vector (linear, polynomial, radial), least absolute shrinkage selection operator, ridge elastic net, partial squares, random forests, recursive partitioning trees, boosted trees. Results show stack performed optimally well predicting seven abovementioned constituents testing set may provide alternative traditional squares method more simultaneous determining properties manure.