摘要
提出了利用R型变量聚类帮助提取体型因子,通过采用主分量、相关系数阵的聚类方法在SAS软件中对原始的人体部位变量进行聚类,从而简化问题。试验结果表明通过R型聚类可以方便地从原始的66个变量提取出8个方面的体型因子来描述人体体型特征。
R-variance clustering was used to extracting body-shape factors in this paper. The body measurements were classified by principle method and correlation matrix in SAS software to simplify the question. The experiment had shown that 8 body-shape factors should be gotten from 66 measurements by variance clustering.
出处
《东华大学学报(自然科学版)》
CAS
CSCD
北大核心
2006年第4期65-69,共5页
Journal of Donghua University(Natural Science)