Customer Validation using Hybrid Logistic Regression and Credit Scoring Model: A Case Study

In this paper a regression model is applied for validating the customers of a company. Using a Delphi method beside the expert panel the main variables which construct the regression model are extracted. A credit scoring system for validation of the customers is developed based on applied regression model. Then a Newton-Raphson method is used for determining the coefficients of regression model. Furthermore the MacFadden statistical value is calculated for approving the regression model. A case study is presented for application of proposed model. Three factors of sponsor's job, applicant's jobs and income are extracted as the main factors which are affecting the customer validation in the case. The result of the proposed model based on the case study showed that using a regression model which is empowered by Delphi system can provide a robust model for validation of the customers for deciding on granting to the customers.