基于混合整数规划法的企业信用风险评估研究  

Evaluation of Enterprise's Credit Risk Based on Mixed Integer Programming Approach

作　　者：薛锋[1] 柯孔林[2] 

机构地区：[1]西安交通大学管理学院,西安710049 [2]浙江工商大学金融学院,杭州310035

出　　处：《中国管理科学》2006年第2期39-44,共6页Chinese Journal of Management Science

基　　金：国家自然科学基金资助项目(70171005)

摘　　要：运用混合整数规划法构建企业信用风险评估模型,并将其应用于我国商业银行信用风险评估中。该模型具有非参数检验特性,不需要样本数据服从正态分布和等协方差,并通过两阶段的分类过程对企业信用状况进行判别。研究结果表明,混合整数规划法有较强的预测贷款企业信用风险的能力。Mixed integer programming approach is used risk in this paper, and is applied to credit risk evaluation to construct the evaluation model of enterprise＇s credit in commercial banks. The model is characterized by the non-parametric test, and don＇t require normal distribution and equality of matrix of covariance. With the method of two - staged classification process, the paper discriminated the credit conditions of enterprises. Empirical study shows that mixed integer programming approach has a higher predicting ability.

关 键 词：混合整数规划法 信用风险评估 LOGIT模型 

分 类 号：F830[经济管理—金融学]