Research of PNN recombining model on personal credit evaluation

被引:0
作者
Jiang, Minghui [1 ]
Chen, Yufang [1 ]
机构
[1] Harbin Inst Technol, Sch Management, Harbin 150001, Peoples R China
来源
DYNAMICS OF CONTINUOUS DISCRETE AND IMPULSIVE SYSTEMS-SERIES B-APPLICATIONS & ALGORITHMS | 2007年 / 14卷
关键词
personal credit evaluation; combining forecast; PNN;
D O I
暂无
中图分类号
O29 [应用数学];
学科分类号
070104 ;
摘要
With the concept of combining forecasts, this paper puts out a new approach by cocombining multi-linear regression and logistic with Elman and BP network. Then, recombine them with, PNN and apply it in personal credit evaluation, of one commercial bank. The results indicate that the combining methods can increase the accuracy, and recombining is a reasonable way for greater precision than either the combining methods or pre-combining models, especially in avoiding recognizing the bad, applications as good ones.
引用
收藏
页码:155 / 158
页数:4
相关论文
共 12 条
[1]  
[Anonymous], COMPUTERS OPERATIONS
[2]  
[Anonymous], J EMPIRICAL FINANCE
[3]   CREDIT SCORING SYSTEMS - A CRITICAL ANALYSIS [J].
CAPON, N .
JOURNAL OF MARKETING, 1982, 46 (02) :82-91
[4]  
ELIZABETH M, 1998, CREDIT RISK MODELING
[5]  
GEXIN GXW, 1994, SYSTEM ENG THEORY PR, V12, P66
[6]  
Guo Z, 2001, LOGISTIC REGRESSION
[7]  
Lewis E., 1992, INTRO CREDIT SCORING
[8]  
LI BX, 2005, J IND ENG MANAGEMENT, V1, P19
[9]  
Shi Q.Y, 2004, STAT RES, V6, P43
[10]   Combined forecasts from linear and nonlinear time series models [J].
Terui, N ;
van Dijk, HK .
INTERNATIONAL JOURNAL OF FORECASTING, 2002, 18 (03) :421-438