Modelling and Forecasting of Aspects of Credit Card Defaults

Abstract

The author of this thesis aims to reproduce and extend the work done by Vladimir G. Drugov [Dru] to understand the dynamics involved in the data set, make conclusions, provide best predictive model to predict future defaults and forecast monthly trends of credits through artificial intelligence. In finance, default is the failure of payment on debt by the due date. This thesis report is devoted to ”modelling and forecasting of aspects of credit card defaults” with the help of Data exploration by statistical visualisation techniques reproduced from The extended part which is research of the author is The data used is that of a credit card company [WEB] which has demo- graphic and financial information of it’s customers and status of default in their credit card payment. The purpose of this study is to: • Find impact of demographic and financial variables on the status of default. • Find important variables responsible for defaults. • Forecast pattern of unpaid credits of the customers.