A Review of Machine Learning Methods For Predicting Churn in the Telecom Sector

Abstract

This paper aims to present an overview of churn, its types, and the machine learning algorithms used to predict it. The paper offers a useful overview of the usage of machine learning algorithms for predicting consumer churn in telecom over the previous five years by examining recent existing models in literature from four scientific databases.. This research indicates that logistic regression is one of the most popular machine learning techniques for churn prediction, which offers high levels of accuracy even with or without data preparation and/or optimization method. The result demonstrates that the studies reviewed used different algorithms and that factors analysis, datasets, and data mining techniques play a big role in determining the algorithm used and its accuracy. In addition, this paper constructs a base for future work to determine the best algorithms to use on datasets, as any proposed model needs to provide logical identification of the determining variables that will cause the customer to churn and have good accuracy when forecasting future churners.

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 — 7 citations exist upstream.

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