Peer-to-Peer (P2P) lending is a Fintech service that allows borrowers of any financial standing to be matched with lenders through online platforms without the intermediation of banks. Correct identification of probable defaulters is important for the longevity of the industry as the lender must bear financial risks should the borrower default, failure of which could result in loss of confidence and pulling out of the platform. However, with more information, it becomes difficult to determine the discriminatory features of the borrower. This study aims to develop a predictive model for loan default prediction in peer-to-peer lending communities. The predictive models were built using Logistic Regression, Random Forest, and Linear SVM with the selected feature set where Random Forest outperformed and achieved an accuracy of 92%. The significant fittest feature subset was obtained using a Genetic Algorithm and was evaluated using a Logistic Regression model. The Random Forest model could be used in the specified domain in this regard in future

Keywords : Genetic Algorithm, Loan Prediction, Peer-ToPeer Lending, Predictive Modelling