Forecasting the Enrolment of Bank Term Deposits: A case study approach with Cross-Industry Standard Process for Data Mining (CRISP-DM) methodology.

Abstract

Abstract This case study employs the Cross-Industry Standard Process for Data Mining (CRISP-DM) methodology to forecast the enrollment of bank term deposits by analyzing a dataset obtained from Moro et al. (2014). The study aims to provide data-driven strategic recommendations to a bank's marketing department by understanding the factors influencing the likelihood of a customer subscribing to a term deposit. The dataset includes various demographic and financial variables, and the analysis employs logistic regression and support vector machine (SVM) models. The CRISP-DM methodology guides the project through stages such as business understanding, data preparation, modeling, and evaluation. The logistic regression model demonstrates high accuracy, and both models meet success criteria, providing valuable insights for enhancing marketing strategies. The study acknowledges limitations, such as the dataset's limited representation of the customer population and potential biases from oversampling, while recommending future research with different machine learning models and extensive datasets for more comprehensive results. The primary objective is to offer strategic guidance to the marketing department, focusing on significant variables and improving communication strategies based on critical findings from the analysis.