This project was built in python to classify loan defaulters. The dataset has about 4 lakh records. Have used Logistic Regression, Decision Trees and Random Forest models for this classification problem and achieved a reasonable accuracy of 0.75 and ROC of 0.65.