FINTECH LENDING CREDIT SCORING MODEL WITH BAYESIAN LOGISTIC REGRESSION APPROACH

• Main Author: VERENA SUDARMO, NETHANIA
• Format: Final Project
• Language: Indonesia
• Subjects: Ilmu alam dan matematika
• Online Access: https://digilib.itb.ac.id/gdl/view/64836
• Institution: Institut Teknologi Bandung

FinTech Lending or Peer-to-peer Lending is a financial service provider that allows lenders and borrowers to meet online. FinTech Lending has been a solution to get easier loan application process for borrowers who do not meet the loan requirements at the Bank. Hence, a credit scoring system is necessary to minimize credit risk. This research is focused on building a model which can predict whether the loan applicants will default or not. The model is expected to maintain the simple characteristics of Fin Tech Lending. The prediction of whether the loan will default or not is based on a Logistic Regression and Bayesian Logistic Regression model with JO variables. ln addition to predicting loan behavior, this research aims to view the effect of each variables to the loan behavior. The result shows no significant difference on the performance from all the models. However, the Bayesian Logistic Regression model with informative prior requires longer duration to compute and perform below the Bayesian Logistic Regression model without informative prior.