Using SVM for Smart Direct Marketing (SDM): A case of predicting bank customers interested in the Term Deposits

Abstract

Abstract:

The objective of this study is to reveal how important and necessary it has become to adopt the new methods and technical of Data Mining derived from artificial intelligence (AI), and this in the process of marketing banking products and precisely that of Term Deposit (TD). Therefore, in our research study we adopted the Support Vector Machine (SVM) method, and we succeeded through the process of modeling borrowed by the techniques of Data Mining (DM) to build a model that will allow us to predict the behavior of bank customers toward term deposits, for this the SVM method has been applied on a database of customers of a bank which includes historical responses of customers during a marketing campaign of the product DT.

For the manipulation tool we used, we chose the Python programming language, recognized by its power in modeling and exploiting DM techniques. For the methodology adopted, we started with a pre-treatment and cleaning of the data to keep only the significant explanatory variables and also for any extreme or aberrant variables, subsequently the modeling was carried out by adopting the DM process, at the end of this process we were able to measure the perfection and the level of predictability of our obtained model, thus we obtained the accuracy of 93% using the metric Accuracy, Thus according to the ROC curve we have that AUC=98, this reflects the performance of our obtained model. So, we were able to build a model that will help bankers make decisions in terms of predicting customers interested in the term deposit product.

JEL Classification : C02, C19, C35, C55, C6, M31,G2

Paper type: Empirical research