+7 (495) 987 43 74 ext. 3304
Join us -              
Рус   |   Eng

Authors

Sorokin S.

Degree
PhD in Physics & Mathematics, Tver State University
E-mail
sergey@tversu.ru
Location
Tver
Articles

Use neural network models in behavioral scoring

Various mathematical and statistical models can be used as basis for constructing scoring systems. These models can be obtained by linear regression, logistic regression, discriminant analysis, decision trees, neural networks, and others. The article describes an example of constructing a model of behavioral skroring in the banking sector, based on a neural network program IBM SPSS Statistics. Also this article presents the technique of modeling, sequence of network settings in the dialogs procedures of the program SPSS, code syntax for the implementation of the considered algorithm, and the simulation of results. These results can be used in the educational process for laboratory work and workshops in teaching methods in neural network modeling program SPSS, as well as independent research. The purpose of writing this article is to review the methodological aspects and step alogoritma constructing a neural network program IBM SPSS Statistics 22 for example, behavioral scoring in the banking sector. Under the behavioral scoring understood mathematical model by which on the basis of a set of known characteristics of the object of study (independent variables) projected a certain desired characteristics (dependent variable), the value of which is not known researcher.
Read more...