MODEL OF CREDIT SCORING AS ALGORITHM OF TYPOLOGY OF FUZZY AGGREGATIONS
V.I. Sonnikova, K.O. Kulidzhoglyan
Keywords: scoring, learning sample, model, classification, logistic regression
Abstract
The paper discusses theoretic and methodic issues of credit scoring model development. The results of the authors research using real date of Novosibirsk bank are presented. Discriminant functions for various groups of clients are formed. The statistical evaluation of reliability of obtained results is made. Conclusions are drawn and recommendations are made.
|
