Abstract In the present study, we have utilized attenuated total reflectance Fourier transform infrared spectroscopy for the examination of raw, pasteurized and adulterated milk samples. Principal component analysis and discriminant analysis have been applied for discrimination and classification purposes. According to the observations and the model of “goodness of fit”, the obtained results explain 100% of the original classification of the dataset and leave one out cross-validation provides 93.74% of accurate classification.

中文翻译：

---

关于生奶、巴氏杀菌奶、奶粉和掺假奶样品的检查及其多元分类：在食品和法医学中的应用

摘要 在本研究中，我们利用衰减全反射傅里叶变换红外光谱来检测生奶、巴氏杀菌奶和掺假奶样品。主成分分析和判别分析已应用于判别和分类目的。根据观察结果和“拟合优度”模型，得到的结果解释了数据集原始分类的 100%，留一交叉验证提供了 93.74% 的准确分类。