K均值和分层聚类法在大米产地鉴别中的应用

Classification of Rice According to the Geographic Origin Based on K-Means Cluster and Hierarchinal Cluster

大米称誉为"五谷之首",是我国的主要粮食作物。大米产地鉴别方法的研究对于保护广大消费者和生产厂家的利益、推动我国大米出口业的发展等都具有重要意义。为此,选取了19个响水大米样品和10个非响水大米样品,使用等离子体原子发射光谱的方法测定了样品中P,B,Zn,Fe,Cu,Mn,Na,K,Mg和Ca等10种微量元素的含量,并利用样品中的微量元素含量数据,使用K-Means聚类法和Hierarchical聚类法对大米样品进行了聚类分析。结果表明,K-Means和Hierarchical聚类法对于样品有聚类趋势。研究结果能够为大米生产厂商与质检部门鉴别大米产地提供参考和借鉴。

Rice is the first of 'the Five Cereals',and is the grain crop of our country.Thus,the classification of the geographic origin of rice is significant for protecting interests of the consumers and producers.Besides,it is important to enhance the quality of rice and develop rice exportation.In this paper,nineteen Xiang-shui rice samples and ten non-Xiang-shui rice samples were analyzed by Inductively Coupled Plasma Atomic Emission Spectrometry to obtain the content value of P,B,Zn,Fe,Cu,Mn,Na,K,Mg and Ca in rice samples,and then classified using chemometrics methods of k-means cluster and hierarchical cluster.The results of K-Means cluster and Hierarchical cluster indicated that two methods both have clustering tendency.A new classification method and technique of rice was provided for producers,sellers and departments of quality and technical supervision in this paper.