Quality grade discrimination of light aroma type liquor based on mass spectrometry and support vector machine

Quality grade discrimination of Chinese liquor was benefit for controlling liquor quality and safeguarding the interests of consumers. In this paper, taking Niulanshan for instance, we studied quality grade discrimination of Chinese liquor with light aroma type. Mass spectra of 57 samples were obtained by head space-solid phase microextraction-mass spectrometry (HS-SPME-MS) technology in the range of m/z 55¹91. And then, the partial least squares regression (PLS) and principal component regression (PCR) models were developed by calibration set and predicted the quality grade of validation set. Obviously PLS model was superior to PCR model. The support vector machine (SVM) models were built by different ion selection methods, PLS regression coefficients and PCR regression coefficients;the prediction accuracy of SVM models for the test set was 80%and 86.7%, respectively. The ions, m/z 71, 103, 104, 106, 127, 149, 161, 179, 183 and 184 were selected by PCR regression coefficients.