摘要
本文在“金翅夜蛾亚科的数值分类研究”的基础上用主成分分析的结果对金翅夜蛾亚科分类的性状做进一步分析,说明各性状对分类的重要性和性状的变异方向,并对用协方差矩阵和相关矩阵进行的主成分分析结果进行比较,说明对本问题(分类指标全是定性指标)适宜用协方差矩阵做主成分分析。
Basing on the research of the numerical taxonomy Plusiinae,the follwing research makes a further analysis on the character of taxonomy of Plusiinae by principal component analysis in order to show the importance of character to taxonomy and the direction of deviation of these characters.It also makes a comparison on the result of principal component analysis which is settled by correlation matrix and covariance matrix so that we can conclude that covariance matrix used as principal component analysis is a suitable method to this research(i.e.taxonomy indices are all qualitative one).
出处
《生物数学学报》
CSCD
北大核心
1996年第4期77-82,共6页
Journal of Biomathematics
关键词
金翅夜蛾亚科
数值分类
主成分分析
协方差矩阵
相关矩阵
Plusiinae,Numerical taxonomy,Principal component analysis,Covariance matrix,Correlation matrix