Year of Graduation
Credit Risk Model Based on Machine Learning Techniques
Applied Mathematics and Information Science
This paper is dedicated to the assessment of borrower’s credit capacity. The review of machine learning methods that classify borrowers into "good" and "bad" is conducted in this study. We investigate SPSS Statistics, KNIME Analytics Platform and R. The models are developed using logistic regression, decision trees and their ensembles. Besides, the model of random forest is created. This model gives the best results and has a good predictive ability.