Comparative Analysis of Logistic Regression and Random Forest Models for Customer Churn Prediction in Laos

Abstract

Customer churn poses a significant challenge in the telecommunications industry, as it directly impacts both revenue and long-term customer retention. This study leverages real-world customer data from TPLUS Digital to evaluate the effectiveness of machine learning techniques in predicting customer churn. The research aims to assess the learning speed and testing capability of machine learning models and compare their performance in churn prediction. Two widely used models Logistic Regression (LR) and Random Forest (RF) were employed and evaluated using various performance metrics, including training and testing time, AUC, Precision, Recall, and F1-Score. The dataset was divided into training and testing sets using multiple ratios: 70%-30%, 80%-20%, and 90%-10%. Results show that Random Forest outperforms Logistic Regression in identifying customers likely to churn and achieves higher overall accuracy. However, Logistic Regression exhibited faster training times and more consistent performance on imbalanced datasets. This study provides valuable insights into the application of machine learning for churn prediction in the telecommunications sector and offers a foundation for developing effective customer retention strategies.