Developed a Machine learning model for loan repayment.
-
Implemented in python programming language.
-
Trained using catboost algorithm.
-
Used heatmaps,boxplots and distribution plots for data understanding.
-
Achieved AUC score of 0.73 on the test set.
-
Achieved overall accuracy of 80% on train and 78 % on the test set.
-
Deployed as a web app using Flask and Heroku .
-
Baseline recall for paying the loan is 77.47%.
-
Baseline recall for not paying the loan is 22.42%.
Link to Project: https://rishabh-third-test-app.herokuapp.com/