Customer Churn Prediction in Telecom Sector with Machine Learning and Information Gain Filter Feature Selection Algorithms

Abstract

Customer churn is a significant issue and one of the primary worries of large businesses. Due to the direct impact on firms’ earnings, particularly in the telecommunications sector, companies are striving to create methods for predicting probable customer churn. Thus, identifying the variables that contribute to client turnover is critical to take the required steps to reduce this churn. Our work’s primary contribution is the development of a churn prediction model that enables telecom carriers to forecast which customers are likely to churn. The model created in this work makes use of machine learning (ML) approaches such as the Support Vector Machine (SVM), the Multi-Layer Perceptron (MLP), the Random Forest (RF), and Naive Bayes (NB). This paper proposes an innovative technique for feature selection that combines the Information Gain and Ranker methods. To evaluate the model’s performance, the accuracy, precision, and F-measure standard measures are used, along with 10-fold cross-validation. The findings gave an accuracy of 95.02% when feature selection is considered and an accuracy of 92.92% without feature selection. The results were compared with the existing methods and our models showed competitive performance in terms of precision and F-measure.

Cite this work

No public copy is stored for this record, so only the metadata is cited.

Stored only when both papers are in the corpus — 27 citations exist upstream.

Papers that cite the same references — the relational fallback while embeddings are still backfilling.