This project is executed within the course "Data Mining I" in HWS18/19 at the University of Mannheim.

The goal of this project is two-fold: Firstly, we want to improve the return on investment for such a tele-marketing campaign by giving the bank a means of reducing the number of clients being called, who are definitely not interested in buying another product, while still calling those customers who are likely to subscribe to a product. In doing so, we support the bank in strengthening customer relations by offering their clients the right product at the right point of time instead of displeasing them with uncontrolled direct marketing campaigns. To achieve the goals stated above, we need to predict whether a customer subscribes to the product or not to eventually decide on a list of customers to call. Since the outcome is either "Subscription - Yes or No", the task is a binary classification problem. To solve it, different classification algorithms are trained and tested on historical data about past phone calls and their success.