线性判别分析
偏最小二乘回归
模式识别(心理学)
支持向量机
人工智能
主成分分析
化学计量学
食品科学
人工神经网络
高光谱成像
判别式
数学
作者
Fernanda Cosme,Juliana Milheiro,Joana Pires,Fernanda Isabel Guerra-Gomes,Luís Filipe-Ribeiro,Fernando M. Nunes
出处
期刊:Food Control
[Elsevier]
日期:2021-07-01
卷期号:125: 107979-
被引量:1
标识
DOI:10.1016/j.foodcont.2021.107979
摘要
Abstract In this study, the feasibility of discriminating monovarietal red wines from Touriga Nacional (TN), Touriga Franca (TF), and Tinta Roriz (TR) from Douro Demarcated Region (DDR) using the anthocyanin profile was investigated in combination with partial least squares – discrimination analysis (PLS-DA), classification and regression trees (CART) and artificial neural networks (ANN). CART and ANN were more accurate in the classification of monovarietal wines according to their origin with a 100% classification rate in the test set. PLS-DA presented an accuracy of 91.4%. The variables most important for monovarietal wines authentication were petunidin-3-glucoside, malvidin-3-acetylglucoside, and cyanidin-3-glucoside for the PLS-DA, petunidin-3-glucoside, peonidin-3-acetylglucoside, and malvidin-3-acetylglucoside for CART analysis, and malvidin-3-acetylglucoside, petunidin-3-glucoside, malvidin-3-glucoside, peonidin-3-coumaroylglucoside and delphinidin-3-glucoside for ANN. The results demonstrate that the anthocyanin profile of Touriga Nacional, Touriga Franca, and Tinta Roriz Douro Denomination of Origin monovarietal red wines combined with multivariate analysis is an effective tool for verifying their authenticity.
科研通智能强力驱动
Strongly Powered by AbleSci AI