The aim of the project is to create a robust machine learning model that predicts the likelihood for a bank's customers to fail on their credit payments for the next month. The dataset used contains information on 24028 customers across 26 variables that includes information regarding whether customer defaulted, credit limits, bill history etc.