Credit Default Prediction System Using Machine Learning

Abstract

The use of innovative technologies and services has allowed financial institutions to offer microcredit to low-income earners. Digitization has allowed lenders to automate loan application components, including underwriting and e-signatures; resulting in more efficient loan delivery while maintaining traditional underwriting and compliance practices. Traditional credit score models have limitations in applying big data technology to build risk models, and machine learning based credit risk models have emerged as a more effective way to predict defaults. This paper proposed a Credit Default Prediction System Using Machine Learning. The scheme successfully designed a classification model to predict loan default before the loan is approved.