谷子主要分量的模糊聚类

主成分分析是现代数量遗传学中对品种资源分析评价的一种常用方法。因为主成分分析可以将一个多变量互相关联的资料系统转换成一组互相独立的变数,每个主成分包括一部分品种资源的遗传信息,而且其重要性可以按照顺序排列起来,这样可以用前面少数几个主成分值来代表亲本资源的基本情况,从而简化工作量。数量性状属“模糊性状”,用模糊聚类法聚类亲本比用其他聚类法要符合实际情况,所以主成分析和模糊聚类法相结合可得到科学的结论。本研究通过主成分分析选出了10个较好品种,并用模糊聚类法将45个谷子亲本聚为7类,可供谷子育种者参考

A FUZZY CLUSTERING ANALYSIS OF PRINCIPAL COMPONETS OF MILLET

The principal component analysis was a freguerty used method in quantitatire genetics as it sinplified the analysis procedure and presented the main genetic information So far the method of fuzzy clustering was the best to be used in practice among clusteving aualysis, On the basis of a three year experinent data, 10 charactevs in 45 varieties of millet were studied with the principal component analysis and 10 varieties were selectel to be excellent; meanwhile, with the method of fuzzy clustering paremts of all varieties tested were grouped into 92 cluster, which was a reference for the millet breeder