Development of a Classification Model for the Prediction of Churn Among Customers Using Decision Tree Algorithm

ABSTRACT

Churn prediction model monitors the customer relationship management in order to preserve the customers who are anticipated to quit from provided service. This study aims to identify variables which are more relevant for the prediction of customer churn which affect the ability to effectively classify customers likely to churn in the future. This study collected secondary data from 7403 customers consisting of information about customer demographics, services subscribed to and account information which was provided by Kaggle online. The study applied the use of two (2) decision tree algorithms namely: CART and C4.5 following which their performance was compared. The results of the study showed that the CART decision tree algorithm created a binary tree with the best accuracy of 81.8% using 95% of dataset for training and 5% of the dataset for testing. The results of the study showed that by increasing the training proportion of the modeling process, the accuracy of the predictive model can be improved. Also, the results showed that CART decision tree algorithm identified a limited yet relevant attributes which were important to the classification of customer churn.

INDEX TERMS Customer churn, Machine learning, Classification, Decision tree