TVS Credit Analytics Case Study
The aim was to predict if a customer will default or not based on customer's loan dataset. Peformed Logistic Regression using glm and lasso both, Random Forest and XGBoost. XGBoost had the most accuracy and AUC and hence was recommended to be used for further predictions.