化学计量学
偏最小二乘回归
指纹(计算)
鉴定(生物学)
模式识别(心理学)
化学
人工智能
生物系统
数学
分析化学(期刊)
计算机科学
色谱法
统计
植物
生物
作者
Huan Fang,Tong Wang,Lan Chen,Xiaozhi Wang,Hai‐Long Wu,Yao Chen,Ru‐Qin Yu
标识
DOI:10.1016/j.jfca.2024.106632
摘要
"Niulankeng Rougui" (NRG) as the top-ranking cultivars in Wuyi Rock tea (WRT) was facing the risk of being faked. Therefore, a chemometrics-assisted excitation-emission matrix fluorescence (EEM) method was proposed for rapid authenticity identification of NRG. To determine the differences between different varieties, 20 tea polyphenols and 1 caffeine in five different WRT were identified by HPLC-DAD, and the contribution of important components to the fluorescence fingerprint of tea was also analyzed. Classification of different WRT (Case 1) and identification of real NRG and adulterated NRG (Case 2) were implemented based on EEM combined with two chemometrics methods, respectively. The correct classification rates (CCRs) were 83.3–100.0 % both in training, test and prediction sets. To predict adulteration levels in NRG, a partial least squares (PLS) regression model was applied and showed good linearity (R2>0.95). The results support that the proposed method is competent for rapid authenticity identification of tea.
科研通智能强力驱动
Strongly Powered by AbleSci AI