基于集对论的信用风险分析

SPA-Based Investigation into Credit Risk

基于集对分析方法及企业的财务数据,建立了信用等级预测模型.通过与判别分析法及Logit分析法的比较,证明该方法能够更好地拟合企业的信用等级,并进行更为准确的预测.研究结果表明,集对分析方法在等级分类和趋势预测这一类似确定又不确定的问题上有良好的应用效果.

Based on set pair analysis (SPA)method and financial data of industrial enterprises,a real world investigation has been carried out by building a forecasting model of the enterprises′ credit rating.Compared with discriminatory analysis and Logistic regression,SPA method demonstrates that in credit risk analysis the forecast results from the new method are more accurate,and the method is thus even more applicable to problems of grade classification and trend forecast.