machine learning for characterization of the phytochemical profile of black tea duringdifferent infusion time

Tea is the second most consumed drink after water, and black tea is the most abundant type of teaproduced worldwide. green and black tea are the most traditional styles of tea productsrepresenting 20–22% and 76–78% of the worldwide production, respectively [1]. green tea comesfrom the unfermented, dried leaves, while black tea is a fully fermented product. the interest intea and the analysis of its chemical compounds has increased significantly due to its nutritional,economic value, and the positive effects of tea on human health [2]. however, few studies haveaddressed the availability of these compounds associated with the infusion time of the black teadrink. for this purpose, machine learning methods were used in this study to investigate the effectof brewing time on its chemical composition. in this way, the following procedure was used: in around-bottom glass flask, 1.0 g of black tea and 20.0 ml of methanol were added, and the mixturewas refluxed at 60°c for 3 h. sampling was done four times between 0 and 3 hours (once everyhour). the gc-fid technique was employed for analysis of the extracts of 100 samples (25samples for each time), and the peak areas were used to construct the machine learning models.various pre-processing methods, such as mean centering (mc), auto-scaling, multiplicativescatter correction (msc), and standard normal variate (snv), were used to remove irrelevantvariations. machine learning methods, including principal component analysis-linear discriminantanalysis (pca-lda), partial least square-discriminant analysis (pls-da), and support vectormachine (svm), were employed to characterize the phytochemical composition of the teasamples. the constructed models showed a distinct class separation between the tea samples thatwere analyzed immediately after adding methanol (0 hour) and the samples corresponding to 3hours sampling. the samples corresponding to brewing time of 1 and 2 hours appear in one cluster.the obtained results showed that the brewing time may therefore be an important factor affectingthe phytochemical composition of black tea.