电子鼻
电子舌
感觉系统
偏最小二乘回归
线性判别分析
感官分析
主成分分析
统计
数学
斯里兰卡
模式识别(心理学)
化学计量学
人工智能
计算机科学
食品科学
心理学
地理
化学
机器学习
认知心理学
环境规划
品味
坦桑尼亚
作者
Zoltán Kovács,István Dalmadi,Larina Lukács,László Sípos,Katalin Szántai‐Kőhegyi,Zoltán Kókai,András Fekete
摘要
Abstract Tea quality used to be traditionally evaluated by sensory panels. Their expertise is considerably better than that of the average consumers. However, there can be definite subjectivity in the judgment of sensory panels. A new direction is advanced sensory study that integrates sensory assessment with instrumental evaluation. The aim of this paper is to distinguish five Sri Lanka black teas from plantations of different geographical origins. From the tea samples, infusions were prepared with standard procedure. Three different test methods were applied: electronic nose, electronic tongue and sensory assessment. In the case of the electronic nose only the volatile components were evaluated. The Alpha Astree II potentiometric electronic tongue analyzed the liquid phase, while in the sensory profiling all perceptible attributes were tested. Data were analyzed with principle component analysis (PCA), linear discriminant analysis (LDA) and in the case of sensory analysis with one‐way ANOVA. The partial least square regression (PLS) was used to estimate the sensory attributes. As a final result, we found that LDA maps show the quality differences of teas grown at different altitudes and geographical location of growing regions. Copyright © 2010 John Wiley & Sons, Ltd.
科研通智能强力驱动
Strongly Powered by AbleSci AI