NORMA eResearch @NCI Library

A Machine Learning Framework to Address Customer Churn Problem Using Uplift Modelling and Prescriptive Analysis

Baskan, Merve (2022) A Machine Learning Framework to Address Customer Churn Problem Using Uplift Modelling and Prescriptive Analysis. Masters thesis, Dublin, National College of Ireland.

[thumbnail of Master of Science]
PDF (Master of Science)
Download (1MB) | Preview
[thumbnail of Configuration manual]
PDF (Configuration manual)
Download (2MB) | Preview


Customer churn refers to the percentage of customers who stop using a product or service in a given time period. Current research uses machine learning and deep learning models to classify customers for that purpose. However, the challenge to conventional customer churn prediction models is that they do not align with the real-life business objective. They only predict the outcome, i.e., whether a customer will churn or not. Models estimating the net effect of customer behaviour like uplift modelling, however, focus on whether a customer is intent on churning and will be retained when targeted with the campaign. This research proposes a machine learning framework to compare the uplift model with the conventional churn prediction model, using predictive and prescriptive analysis. The framework presents XGBoost(eXtreme Gradient Boosting) and Logistic Regression churn prediction models and their uplift models with two different treatments. Results of the four models are evaluated in this paper based on Qini coefficient, Qini curve, treatment correlation and accuracy. The results show that the uplift model outperforms the conventional customer churn prediction model, when it comes to targeting the right customer group for a retention strategy.

Item Type: Thesis (Masters)
Uncontrolled Keywords: uplift modelling; customer churn prediction; marketing
Subjects: Q Science > QA Mathematics > Electronic computers. Computer science
T Technology > T Technology (General) > Information Technology > Electronic computers. Computer science
H Social Sciences > HF Commerce > Marketing > Consumer Behaviour
Q Science > Q Science (General) > Self-organizing systems. Conscious automata > Machine learning
Divisions: School of Computing > Master of Science in Data Analytics
Depositing User: Tamara Malone
Date Deposited: 18 Jan 2023 15:59
Last Modified: 06 Mar 2023 17:03

Actions (login required)

View Item View Item