A novel freezing crystallization‐HPLC method combined with machine learning for determining pigments and geographical classification of extra virgin olive oil

• DOI: 10.1002/aocs.12947
• Publication Date: 2025-03-11T04:15:31Z
• Authors: Cong‐Hui Lu; Yu Gao; Hui‐Yuan Lu; Wei‐Jian Shen; Jules Muhire; Zhi‐Bin Lu; Quan Jing; Xin‐Yi Huang; Dong Pei; Duo‐Long Di

AbstractEffective removal of the fatty acid matrix and enrichment of trace target components is a key step in the quantitative analysis of minor components in edible oils. In this study, a novel sample pretreatment method named freezing crystallization was developed to analyze pigments in extra virgin olive oil (EVOO). The limits of detection and limits of quantification of this method were 0.125–0.625 μg/mL and 0.5–2.5 μg/mL, respectively. Linear correlations were obtained (r2 ≥ 0.9995), and the recoveries at three spiked levels were 84.2%–105.8%. Besides, the primary pigment components information combined with machine learning to classify the origin of Chinese EVOOs. The k‐nearest neighbor (kNN), decision tree (DT), and random forest (RF) were employed to classify the origin of EVOOs, and the accuracies were up to 88%, 88%, and 96%, respectively. This result shows that the novel method has good accuracy and stability, and pigments can be used as a basis for classifying the geographical origin of Chinese domestic EVOOs.