Comparison of Analytical Ability of PLS and SVM Algorithm in Estimation of Moisture Content, Higher Heating Value, and Lower Heating Value of Cassava Rhizome Ground using FT-NIR Spectroscopy

Abstract

FT-NIR spectroscopy coupled with chemometrics analysis was used for nondestructive estimation of moisture content (MC), higher heating value (HHV) and lower heating value (LHV) of cassava rhizome ground. The goal of this study was compared to the analytical ability of both algorithm between PLS and SVM. The purpose was to find the effective modelling technique. The outcome was found that PLS and SVM provided good accuracy in evaluation of energy properties, and could be utilized for quality assurance. PLS algorithm gave slightly higher accuracy than SVM algorithm for the prediction of MC, HHV, and LHV. PLS regression generated no difference between measured and predicted value. PLS and SVM regression showed R2 between 0.90-0.98 and 0.84-0.90 for all parameters, respectively. The pre-processing of 2nd derivative was suitable for the PLS and SVM regression to the modelling.