Abstract: Fenjiu, which was a reprehensive of fragrant liquor, was taken as the research object. The pH values, electrical conductivities and the contents of flavor components analyzed by GC-MS were systematically analyzed with the age of liquor. Based on this, with the help of multiple linear regression (MLR) and partial least square regression (PLSR), a model of liquor property and its age was established. The estimation stability and generalization ability of the models were strictly analyzed by both internal and external validation. The correlation coefficient R of established models, leave-one-out (LOO) cross-validation (CV), Qext of MLR were 0.9990, 0.9980, 0.9984; the result of PLSR was the following: R and RLOO were 0.9949, 0.9840. The results showed that the model is of great robustness and good predictive abilities. The method may play an important role in identifying the Fen-flavor liquor age in the theoretical and application.

Key words: liquor, identification of liquor age, gas chromatography-mass spectrometry (GC-MS)