id	author	title	date	pages	extension	mime	words	sentence	flesch	summary	cache	txt
ajst-26214	Fu, Ni; Tuo, Xianguo; Zhang, Wei	Grade Identification of Strong-flavor Raw Liquor Based on GC-MS Combined with Spearman-KPCA Feature Extraction	2024	10	.pdf	application/pdf	5610	240	49	This is because after Spearman correlation analysis, the characteristic compounds that have a greater influence on the classification of raw wine grades are retained, so that the principal components after secondary feature extraction can better characterize the change of the quality of raw wine; the prediction accuracy of the model established by Spearman-KPCA combined with SVM, XGBoost and BP neural network is higher than that of Spearman-PCA, and this may be due to the fact that the change of the content of the substances in the raw wines of different grades is not simple linear increase or decrease. [12] applied GC- MS spectrum technology to the identification of Xiaoqu pure flavor raw wine of different grades, and established a grade discrimination model of Xiaoqu pure flavor raw wine in combination with PL-DA to achieve accurate identification of raw wine samples.	cache/ajst-26214.pdf	txt/ajst-26214.txt