摘要
以工程造价咨询企业为研究对象,构建其信用评价指标体系;针对其信用信息数据搜集难度大、指标数量多、相关性强的特点,以主成分分析法作为指标约简和赋权方法,以三角白化权函数灰色聚类作为信用分类方法构建其信用评价模型;最后以宁波市30个工程造价咨询企业为样本集进行实证分析,结果表明该评价模型能够较好地处理小样本、高维数的信用数据,同时具有准确率高、分类区分度好、误判成本低的优点。
With cost engineering consuhing enterprises as the research object, the paper establishes a credit evaluation in- dex system. The basic data of credit information is difficult to collect, meanwhile, the correlations between the numerous indexes are relatively strong. To solve this problem, a credit evaluation model of cost engineering consulting enterprises is established by using principal component analysis to reduce the index dimensions and the grey triangle whitenization weight function clustering method to classify their credit level. In addition, the model is tested by 30 cases of cost engineering con- suiting enterprisesin the city of Ningbo. The empirical research results indicate that the established model is able to deal with the small sample and high dimension with good forecast ability, high discriminability and low misjudge costs.
出处
《科技管理研究》
CSSCI
北大核心
2014年第22期166-171,共6页
Science and Technology Management Research
基金
国家自然科学基金项目"建筑市场执业资格人员信用行为及其治理模型研究"(71172148)
关键词
信用评价
工程造价咨询企业
灰色聚类
主成分分析
credit evaluation
cost engineering consulting enterprise
grey clustering
PCA