主成分分析法用于食品样品分类研究

Classification of food samples by principal component analysis

利用化学模式识别技术的主成分分析(PCA)法对取自江西不同地区的谷物、蔬菜、水果的24个样品进行分类研究,所涉及的变量包括金属离子K、Na、Ca、Mg、Mn、Fe、Zn、Cu、Cr、Pb和有机物蛋白质、碳水化合物、脂肪、VB1、VB2。结果表明,运用PCA法可获得大量样品的统计特征,对变量进行适当的选择能够有效地分辨不同的样品。

Pattern recognition technique was applied to classify the food samples from different area in Jiangxi province. In this study the technique of principal component analysis (PCA) of pattern recognition to distinguish the grain, vegetable and fruit samples was adopted. The variables included the metal elements of K, Na, Ca, Mg, Mn, Fe, Zn, Cu, Cr and Pb and the organic compounds of protein,carbohydrate, fat, VB1 and VB2. The optimized classification is based on well selecting the variables. The results show that PCA technique may offer the potential strategy in the distinguishing classes of the food samples.