摘要
在考虑企业可能对外发布一些虚假信息这一前提下,借鉴DBSCAN聚类算法的思想,研究了圆形邻域的孤立点挖掘算法,并将其应用于企业虚假信息的识别,在此基础上提出并定义了企业信用风险失真度的概念及其量化结构,以此分析虚假信息对企业信用风险造成的影响,为投资者决策提供理论依据.
Under the assumption that the enterprise maybe issues some false information, this paper discusses the circular neighborhood outlier mining algorithm by extracting the idea of DBSCAN clustering algorithm, and applies the algorithm to identify the false information. Moreover, the paper proposes the concept and quantified structure of credit risk distortion. Using this, the paper analyzes the impact on the credit risk of the false information, and the conclusions can provide some rational basis for investors.
出处
《数学的实践与认识》
CSCD
北大核心
2012年第4期94-101,共8页
Mathematics in Practice and Theory
基金
教育部人文社会科学研究项目(10YJC630334)
山东省自然科学基金(2009ZRB019AV
ZR2009HL002)
关键词
虚假信息
信用风险失真度
孤立点挖掘
向量空间
false information
enterprise credit risk distortion
outlier mining
vector space