Share
Export Citation
Increasing Accuracy of Ensemble Logistics Regression Classifier by Estimating the Newton Raphson Parameter in Credit Scoring
Aziz F.
5th International Conference on Computing Engineering and Design Icced 2019
Abstract
The large volume of customer data in the credit the industry makes the development of an effective credit scoring model extremely important. The use of an ensemble model on statistical methods to solve credit scoring problems managed to get the best predictive performance. Ensemble performance can still be improved by estimating the parameters using nonlinear equations. This paper proposes the estimation of the Ensemble Logistics Regression using Newton Raphson parameter in order to increase the accuracy of the ordinary Logistics Regression model. The results showed that our proposed method successfully improved the accuracy performance of the origin classifier up to 2.6% in the credit scoring.