Rapid Identification of Adulterated Camellia Oil Based on Mass Spectrometry Features and Logistic Regression

To achieve the rapid identification of adulterated camellia oil, this paper proposes a mass spectrometry-based method. The Pearson correlation coefficient method and principal component analysis are used to extract the mass spectrometry features, and combined with logistic regression to establish an adulteration identification model for camellia oil adulterated with 20% rapeseed oil, rice bran oil, peanut oil, and sesame oil. The leave-one-out cross-validation and an independent test set are used for double verification to evaluate the generalization ability of the model. The experimental results show that the accuracy of the model in the leave-one-out cross-validation reaches 96.10% (F1 value: 97.30%), and the accuracy on the independent test set is 94.12% (F1 value: 94.00%). The difference between the two types of verification results is less than 2%, indicating that the model has good generalization and stability under different data distributions. Compared with KNN, decision tree, and SVM, logistic regression achieved higher accuracy rates by 8.83, 2.94, and 2.94 percentage points, respectively. This study provides reliable technical support for the quality and safety control of camellia oil. In the future, the robustness of the model can be further improved through data expansion or regularization optimization.