This invention discloses a method for identifying
coconut oil based on MALDI-TOF
mass spectrometry and a committee strategy, belonging to the field of
analytical chemistry and
food safety testing technology. The method includes: acquiring and preprocessing
mass spectrometry data of the oil sample; extracting several characteristic peaks and their corresponding relative intensities from the
mass spectrometry data; calculating the
score of the oil sample based on the mass-to-charge ratio and intensity of the characteristic peaks; and identifying whether the oil sample is
coconut oil based on its
score. This invention optimizes the characteristic peak combination for
coconut oil authenticity identification based on
mass spectrometry combined with a committee strategy. The constructed
machine learning classification model achieves 100% accuracy and an AUC value of 1 on the
test set, solving the problems of incomplete
feature extraction,
instability of
mass spectrometry signals, and peak drift interference in existing
machine learning models, thus achieving accurate identification of edible oils. This invention has a wide range of applications and can be extended to the identification of other types of edible oils, exhibiting strong versatility and comprehensively covering the identification needs of the
edible oil market.