A comparison of multi-classification methods for credit risk assessment in commercial banks

被引：0

作者：

Zhou Difeng ^{[1
]}

Lin Chengde ^{[1
]}

机构：

[1] Xiamen Univ, Automat Dept, Xiamen 361005, Peoples R China

来源：

Proceedings of the 24th Chinese Control Conference, Vols 1 and 2 | 2005年

关键词：

Credit Risk Assessment; one-against-rest strategy; One-against-One strategy; Directed Acyclic Graph(DGA) strategy; embedded space strategy;

D O I：

暂无

中图分类号：

TP [自动化技术、计算机技术];

学科分类号：

0812 ;

摘要：

Credit Risk Assessment problem belongs essentially to an ordinal classification problem. How to construct an available assessing model is an important part in commercial banks credit management. In this paper, several multi-classification strategies, such as "One-against-rest", "One-against-one", "Directed Acyclic Graph", and "Embedded Space", are compared and applied them to construct credit risk assessment models respectively. By experiments, their assessing performances are compared and DAGSVM and ESSVM are more suitable for our purpose.

引用

页码：1734 / 1737

页数：4