24th Conference of the International Federation of Operational Research Societies
Abstract Submission

299. Discovering specific customers using Machine Learning for Marketing Strategies.

Invited abstract in session Ethics and OR, Public Service and Societal Complexity, stream OR and Ethics, and Societal Perspectives.

Authors (first author is the speaker)

1. Ioannis Tsiligkaridis
Heritage University

Abstract

Our business model is based on Machine Learning (ML) techniques that can attract more customers to buy bank products. We focus on how much each feature of a dataset contributes to the model predicting customer behavior.
An ensemble model consisting of Random Forest and XGBoost classifiers is used to identify potential customers. The XGBoost is strongly reliant on a certain variable.
The feature importance of both ensembles uses socioeconomic indicator variables from a bank’s demographic attributes dataset, and it serves as a global explainability method for business decision making.

Keywords

Status: accepted


Back to the list of papers