Abstract
This study assessed the feasibility on simultaneous and rapid measurement of total polyphenol and caffeine contents in black tea infusion using a developed spectroscopy system. First, a portable visible/near infrared (VIS-NIR) spectroscopy system was developed for data acquisition, and then synergy interval partial least squares (Si-PLS), backpropagation artificial neural network (BP-ANN), and adaptive boosting (AdaBoost) combined with BP-ANN, namely BP-AdaBoost, were used comparatively for modeling. The performances of these models were evaluated according to the correlation coefficient (R p) in the prediction set. Experimental results showed that the performances of nonlinear BP-ANN and BP-AdaBoost models were superior to those of linear Si-PLS models. In particular, BP-AdaBoost models have made a great progress in contrast to BP-ANN models, with R p = 0.9382 for total polyphenol content and R p = 0.9471 for caffeine content. The overall results indicated that this developed spectroscopy system combined with a suitable multivariate calibration tool could be used for the simultaneous and rapid measurement of the main compositions in black tea infusion.
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have