基于支持向量机回归集成的小微企业信用风险度评估模型研究  

Study on the Micro and Small Enterprises Credit Risks Evaluation Model Based on Support-vector Machine Regression Ensemble

作　　者：夏晗 Xia Han(School of Business Administration, Zhongnan University of Economics and Law, Wuhan 430073,Hubei,China;School of Economics and Management, Huanghuai University, Zhumadian 463000, Henan,China)

机构地区：[1]中南财经政法大学工商管理学院,湖北武汉430073 [2]黄淮学院经济与管理学院,河南驻马店463000

出　　处：《征信》2019年第4期21-27,共7页Credit Reference

基　　金：河南省软科学研究计划项目(182400410155);河南省高等学校重点科研项目(18A630038);河南省高等学校人文社会科学研究项目(2018-22JJ-328)

摘　　要：小微企业信用风险评估体系的不完善导致小微企业融资难和贷款违约率高等问题。设计包括企业特质、企业财务指标、企业主特质和贷款方式在内的小微企业信用风险评价指标体系,利用具有小样本学习优势的模糊积分支持向量机回归集成方法,构建小微企业信用风险度评估模型,并将此模型与支持向量机、神经网络等模型对比。实证结果表明该模型具有较高的精度和效率,证实了模型的可行性和优越性,为小微企业信用风险评估系统的构建提供了依据。The imperfectness of the micro and small enterprises (MSEs) credit risks evaluation system leads to financing difficulties for MSEs and their high default ratio on loans, etc. This article designs a kind of MSEs credit risks evaluation index system which includes the characteristics of the enterprises, the financial indexes of the enterprises,the characteristics of the enterprises’owners and their pattern of lending, constructs a kind of MSEs credit risk evaluation model by making use of fuzzy integral support-vector machine regression ensemble method with the advantage of small sample learning, and compares this model with the support-vector machine model and the neural network model, etc. The empirical study demonstrates that this model has relatively high accuracy and efficiency and verifies its feasibility and advantage, providing a basis for the construction of MSEs credit risks evaluation system.

关 键 词：支持向量机集成 信用风险评估 小微企业 模糊积分 

分 类 号：F832.33[经济管理—金融学] F276.3