PCA和PLS-DA用于晒青毛茶级别分类研究

利用主成分分析法(Principal Component Analysis,PCA)和偏最小二乘法判别分析(Partial Least Squares Discriminant Analysis,PLS-DA)对晒青毛茶进行级别分类,并通过统计分析找出重要理化成分。结果表明:PCA和PLS-DA均可以直观地对晒青毛茶级别进行分类,其中能够稳定地分类出3级毛茶,而难以将6级和9级毛茶明显地分类。通过PCA载荷图(Loadings plot)和PLS-DA变量重要性因子(Variable important for the projection,VIP)分布图可得出氨基酸含量为级别分类的重要理化成分,其中赖氨酸(Lys)、脯氨酸(Pro)和苯丙氨酸(Phe)是对级别分类最重要的3种氨基酸组分。

Study of Pu′er Raw Materials Grade Classification by PCA and PLS-DA

Two classification methods for Pu′er raw materials were explored using principal component analysis (PCA) and partial least squares discriminant analysis (PLS-DA), and the important physical and chemical compositions were identified through the statistical analysis. The results revealed that both PCA and PLS-DA could directly classify the grades of Pu′er raw materials, particularly for the grade 3, but not for that of grade 6 and 9. The PCA loadings plot and PLS-DA variable important for the projection plot indicated that the contents of amino acids were the important physical and chemical components for classification. Lysine (Lys), proline (Pro) and phenylalanine (Phe) were three most important physical and chemical compositions.