Comparing the analytical performances of Micro-NIR and FT-NIR spectrometers in the evaluation of acerola fruit quality, using PLS and SVM regression algorithms
Accuracy in term of prediction ability of a state-of-the-art portable NIR is discussed. For the first time, acerola (Malpighia emarginata DC.) is studied by NIR spectroscopy. Titratable acidity and ascorbic acid are modelled with the NIR handle-held device. A linear (PLS) and a non-linear (SVM) regression algorithm are critically compared. Passing-Balblok regression compared the two devices and them to the reference data.