二次分类器
线性判别分析
支持向量机
灵敏度(控制系统)
人工智能
离心
模式识别(心理学)
色谱法
分析化学(期刊)
数学
化学
生物系统
计算机科学
生物
工程类
电子工程
作者
Gema Puertas,Patricia Cazón,Manuel Vázquez Vázquez
出处
期刊:Food Chemistry
[Elsevier]
日期:2023-02-01
卷期号:402: 134507-134507
被引量:6
标识
DOI:10.1016/j.foodchem.2022.134507
摘要
The aim of this work was to develop a quick and cheap method for fraud detection in egg labels according to the four legal farming method of the EU. The plasma obtained from egg centrifugation was investigated for this purpose. Initial protein content in egg, plasma protein content, plasma colour parameters (L*, a* and b*) and plasma UV-VIS-NIR (Ultraviolet-Visible-Near-infrared) spectra were evaluated. The classification algorithms applied were SVM (Support-Vector-Machine), LDA (Linear-Discriminant-Analysis) and QDA (Quadratic-Discriminant-Analysis). The analysis of the protein content did not detect differences. Colour parameters and spectral measurements showed significant differences between eggs. Spectra analysis with QDA gave sensitivity of 100% in the calibration set. The validation set scored 87.5% sensitivity and 94.07% specificity using the visible spectra. This work established plasma spectral measurements combined with classification algorithms as a powerful tool to discriminate the four farming systems. This work presents a fast tool for the egg label control.
科研通智能强力驱动
Strongly Powered by AbleSci AI