Lending club default prediction using Naïve Bayes and decision tree

Abstract

Currently P2P lending is one of the most emerging disruptors in the financial sector. Lending Club is a P2P platform based in America. Besides its flexibility to give instant lending this industry have high risk for their investors to lending money. In order to mitigate this risk, this study aims to predict the default risk using decision tree J48 and naive bayes. One of the results in this research show that J48 and Naïve Bayes are both good in predicting the default in P2P lending sector. Another contribution of the paper might be useful for similar companies to see which factors that influence the most to default loan status.