基于投影寻踪技术的供应商评价与实证研究  

Supplier Evaluation and Empirical Research Based on Projection Pursuit Technology

作　　者：于晓虹 王芊月 吴芹 YU Xiao-hong;WANG Qian-yue;WU Qin(College of Humanities and Law,Shanghai Business School,Shanghai 200235,China)

机构地区：[1]上海商学院文法学院,上海200235

出　　处：《价值工程》2023年第23期53-55,共3页Value Engineering

基　　金：2021年度上海商学院校级一流本科课程建设项目——市场调查与统计(项目编号:SBS-2021-XJJK-17)。

摘　　要：研究根据已有的供应商选择与评价指标体系和30个供应商样本数据,建立了供应商评价与选择的PPC模型。结果表明:在21个评价指标中,降低成本计划(指标x5)对供应商能力的影响最大,其次是总资产周转率(指标x14),产品退货比率(x2)的影响最小;在30个供应商中,供应商26的竞争力最强,其次是供应商25,第30号供应商最劣;供应商可以分为五类,类内供应商差异很小,而不同类之间差异显著。采用PPC建模,能同时构建每个样本得分的评价函数和得到每个评价指标的最佳客观权重,完成评价指标和供应商的优劣排序和分类,同时准确揭示不同评价指标数据之间的正负相关关系。因此,在供应商选择与评价中应优先推荐使用PPC模型。Based on the existing supplier selection and evaluation index system and 30 supplier sample data,a PPC model for supplier evaluation and selection was established.The results show that among the 21 evaluation indicators,the cost reduction plan(indicator x5)has the greatest impact on supplier capability,followed by the total asset turnover rate(indicator x14),and the product return rate(indicator x2)has the smallest impact.Among the 30 suppliers,supplier 26 has the strongest competitiveness,followed by supplier 25,and supplier 30 is the worst;Suppliers can be divided into five categories,with small differences within each category,while significant differences between different categories.Using PPC modeling,it is possible to simultaneously construct an evaluation function for each sample score and obtain the optimal objective weight for each evaluation indicator,complete the ranking and classification of evaluation indicators and suppliers,and accurately reveal the positive and negative correlation between different evaluation indicator data.Therefore,the PPC model should be prioritized in supplier selection and evaluation.

关 键 词：供应商 评价与选择 投影寻踪 评价指标 权重 

分 类 号：TP278[自动化与计算机技术—检测技术与自动化装置]