Applying Data Mining Technology for Customer Lifetime Value (CLV) Prediction: the Case of Commercial Bank of Ethiopia (CBE)
No Thumbnail Available
Date
2023-10
Authors
Journal Title
Journal ISSN
Volume Title
Publisher
Addis Ababa University
Abstract
Data mining is a powerful tool for businesses to discover hidden patterns and insights from their data. This can be used to improve customer relationship management (CRM), particularly in understanding the key factors that contribute to predicting customer lifetime value (CLV). CLV is the present value of all future profits that a customer will generate throughout their relationship with a company. One of the various service sectors that gathers, manages, and retains enormous volumes of data over time is the Commercial Bank of Ethiopia (CBE). This data can be used to predict CLV and improve CRM by understanding the key factors that influence customer behavior.
This study used data mining techniques to predict CLV at CBE using the Cross-Industry Standard Process for Data Mining (CRISP-DM). The dataset included information on 100,096 customers and 19 attributes, such as demographics, account details, usage, and transactions. After business understanding and data understanding the data was prepared for experimentation. This involved removing outliers and converting categorical variables to numerical values. Based on their benefits and prior performance reported in the literature, the three machine learning algorithms linear regression, random forest, and decision tree were chosen for the experiment. The algorithms were implemented and their performance was assessed using the Python programming language. Using R2 and RMSE, the models' performance was assessed.
The results revealed that random forest regression had the highest R2, at 86.8%, followed by decision tree regression at 72.4% and linear regression at 56.91%. The study also found that the most important features for CLV prediction are transaction-related features, such as debit amount, credit amount, and total number of transactions. This study demonstrates the applicability of data mining techniques to improve customer relationship management at CBE. By understanding the key factors that contribute to CLV, CBE can develop targeted interventions to keep its customers engaged and grow its business.
Description
Keywords
Data Mining, Machine Learning Algorithms, Customer Lifetime Value