基于主成分分析法的供应商评价指标筛选

Slection of Supplier Evaluation Index Based on Principal Component Analysis

针对传统的供应商评价与选择方法的主观性过强的问题,文章利用相似性度量理论中的x2统计量,在专家评分的基础上,构造出每两两指标之间的相关系数矩阵,再运用主成分分析法做因子分析,根据因子载荷矩阵得出主成分与原始指标的相关系数,再确定出阈值,剔除相关系数小于阈值的指标,从而实现指标的筛选。最后,以L公司的供应商为例,进行了实证研究。验证了该方法的实用性和有效性。

In view of the fact that the traditional methods of supplier evaluation and selection are too subjective. Based on the chi-square statistic in similarity measure theory, the corre- lation coefficient matrix between indexes is constructed on the basis of expert scores, then the principal component analysis is used to do factor analysis. According to the factor load ma- trix, the correlation coefficient between the principal compo- nent and the original index is obtained, and then the thresh- old is determined, and eliminate the index of correlation co- efficient which is smaller than the threshold. So as to filter the index. Finally, taking a company as an example, an em- pirical study is conducted. The practicability and effectiveness of the method are verified.