| 基于变量双重检验的Fisher信用风险度量模型 |
| |
| 引用本文： | 何树红,王善民.基于变量双重检验的Fisher信用风险度量模型[J].系统工程,2007,25(8):15-20. |
| |
| 作者姓名： | 何树红 王善民 |
| |
| 作者单位： | 云南大学,数学系,云南,昆明,650091 |
| |
| 基金项目： | 国家自然科学基金;云南大学专项经费资助项目 |
| |
| 摘 要： | 为提高模型入选样本变量的准确性与稳定性，本文采用附加信息检验和多重共线性检验选择自变量，并应用Fisher判别原理建立信用风险度量函数。实证结果显示，所建立的模型对全部246个样本的误判率仅为6．91％，准确率为93．09％；而对风险样本的误判率为4．55％，正确辨别精度高达95．45％。
|
| 关 键 词： | 信用风险 附加信息检验 多重共线性检验 Fisher判别 |
| 文章编号： | 1001-4098（2007）08-0015-06 |
| 修稿时间： | 2006-12-15 |
Fisher Discriminant Model of Credit Risk Based on Double Test to the Independent Variables |
| |
| HE Shu-hong,WANG Shan-min.Fisher Discriminant Model of Credit Risk Based on Double Test to the Independent Variables[J].Systems Engineering,2007,25(8):15-20. |
| |
| Authors: | HE Shu-hong WANG Shan-min |
| |
| Institution: | Department of Mathmeties,Yunnan University,Kunming 650091 ,China |
| |
| Abstract: | Applying additional test and multilinear test, independent variables are selected, and then Fisher diseriminant function and credit-risk model are established. The result shows that, to all samples, the error rate of the model discriminent is 6.91 % ,while the accuracy rate is 93.09%. As to venture sample,the error rate is merely 4.55% ,and the accuracy is up to 95.45%. |
| |
| Keywords: | Credit Risk Additional Test Multilinear Test Fisher Discriminant |
| 本文献已被 维普 万方数据 等数据库收录！ |
|
