主成分分析
高效液相色谱法
线性判别分析
偏最小二乘回归
色谱法
化学计量学
化学
模式识别(心理学)
人工智能
数学
统计
计算机科学
作者
Guillem Campmajó,Laura Cayero,Javier Saurina,Óscar Núñez
出处
期刊:Foods
[Multidisciplinary Digital Publishing Institute]
日期:2019-08-01
卷期号:8 (8): 310-310
被引量:15
摘要
Hen eggs are classified into four groups according to their production method: Organic, free-range, barn, or caged. It is known that a fraudulent practice is the misrepresentation of a high-quality egg with a lower one. In this work, high-performance liquid chromatography with ultraviolet detection (HPLC-UV) fingerprints were proposed as a source of potential chemical descriptors to achieve the classification of hen eggs according to their labelled type. A reversed-phase separation was optimized to obtain discriminant enough chromatographic fingerprints, which were subsequently processed by means of principal component analysis (PCA) and partial least squares-discriminant analysis (PLS-DA). Particular trends were observed for organic and caged hen eggs by PCA and, as expected, these groupings were improved by PLS-DA. The applicability of the method to distinguish egg manufacturer and size was also studied by PLS-DA, observing variations in the HPLC-UV fingerprints in both cases. Moreover, the classification of higher class eggs, in front of any other with one lower, and hence cheaper, was studied by building paired PLS-DA models, reaching a classification rate of at least 82.6% (100% for organic vs. non-organic hen eggs) and demonstrating the suitability of the proposed method.
科研通智能强力驱动
Strongly Powered by AbleSci AI