One of the most important issues in the field of quality assurance of olive oils is the detection of the freshness of olive oil. In this study, 400 nm laser-induced fluorescence spectroscopy was used with supervised and unsupervised multivariate analysis methods to develop a rapid method able to discriminate between freshly produced olive oils and oil that has been stored for a period of time ranging from 12 to 24 months. The fluorescence spectral data were firstly processed by the PCA. This method shows strong discrimination of the three oil classes using the first three components which present 96% of the total variability of the initial data, and then supervised classification models were constructed using the discriminant partial least square regression PLS-DA, support vector machine SVM, and linear discriminant analysis LDA. These methods show a high capacity in the classification of the three classes of olive oil. The validation of these classification models by external samples shows a high capacity of classification of the samples in their class with an accuracy of 100%. This study demonstrated the feasibility of the fluorescence spectroscopy fingerprint (routine technique) for the classification of olive oils according to their freshness and storage time.

中文翻译：

---

使用荧光光谱和化学计量算法快速分析橄榄油的新鲜度的方法。

橄榄油质量保证领域中最重要的问题之一是检测橄榄油的新鲜度。在这项研究中，将400 nm激光诱导的荧光光谱与有监督和无监督的多变量分析方法一起使用，以开发一种能够区分新鲜生产的橄榄油和已储存12到24个月的橄榄油的快速方法。 。荧光光谱数据首先由PCA处理。该方法显示了使用前三个成分（代表初始数据的总变异性的96％）对三种油类的强烈区分，然后使用判别式偏最小二乘回归PLS-DA，支持向量机SVM构建了监督分类模型，以及线性判别分析LDA。这些方法在对三类橄榄油进行分类中显示出很高的能力。通过外部样本对这些分类模型进行的验证显示出其类别中样本的分类能力很高，准确性为100％。这项研究证明了荧光光谱指纹图谱（常规技术）根据橄榄油的新鲜度和储存时间对橄榄油进行分类的可行性。