Quantification of cow milk in adulterated goat milk using Raman spectroscopy and machine learning

文献类型： 外文期刊

第一作者： Zhang, Yinsheng

作者： Zhang, Yinsheng;Shen, Binge;Wang, Haiyan;Zhao, Yaju;Zhang, Yinsheng;Wang, Haiyan

作者机构：

关键词： Goat milk; Cow milk; Adulteration; Raman spectroscopy; Regression model

期刊名称：MICROCHEMICAL JOURNAL （ 影响因子：5.1； 五年影响因子：4.7 ）

ISSN： 0026-265X

年卷期： 2025 年 215 卷

页码：

收录情况： SCI

摘要： Goat milk is frequently adulterated with low-cost milk alternatives. This study presents a method that integrates Raman spectroscopy and Gaussian-Weighted k-Nearest Neighbor Regression (Gk-NNR) for quantifying cow milk in adulterated goat milk. Goat milk samples were adulterated with varying amounts of cow milk and analyzed using Raman spectroscopy. A Gk-NNR prediction model was developed to estimate the level of adulteration, utilizing a stratified Kennard-Stone split method for balanced data division. This model extended the traditional k-NNR by assigning Gaussian-like weights within the k-neighbor vicinity. It achieved a mean squared error (MSE) value of 0.005 and an R2 value of 0.962, surpassing peer methods such as linear regression, ridge regression, least absolute shrinkage and selection operator, partial least squares regression, random forest regression, support vector regression, and k-NNR, which had MSE values of 0.080, 0.024, 0.024, 0.024, 0.032, 0.012, 0.007, respectively and had R2 values of 0.332, 0.802, 0.801, 0.800, 0.730, 0.896, 0.939, respectively. These findings demonstrate that combining Raman spectroscopy with Gk-NNR provides a rapid and effective technique for detecting goat milk adulteration.

分类号：

  • 相关文献
作者其他论文 更多>>

微信小程序