A Novel and Robust LSTM Model for Customer Churn Analysis Using Deep, Machine Learning, and Ensemble Learning: A Telecommunications Case
| dc.authorid | https://orcid.org/0000-0002-7996-9169 | |
| dc.authorid | https://orcid.org/0000-0001-5896-8880 | |
| dc.authorid | https://orcid.org/0000-0002-8752-3335 | |
| dc.contributor.author | Başarslan, Muhammet Sinan | |
| dc.contributor.author | Ünal, Aslıhan | |
| dc.contributor.author | Kayaalp, Fatih | |
| dc.date.accessioned | 2025-08-11T12:48:53Z | |
| dc.date.available | 2025-08-11T12:48:53Z | |
| dc.date.issued | 2025 | |
| dc.department | İktisadi İdari ve Sosyal Bilimler Fakültesi | |
| dc.description.abstract | Customer churn is an important issue in increasing both the long- and short-term revenues. If companies identify customers’ churn behavior, they can prevent churn, ensure customer loyalty, and, in turn, gain better financial returns. The telecommunications sector is a customer-oriented sector that requires customer retention to survive in the market. In this sector, customer churn is observed at a high level. In recent years, artificial intelligence-based customer churn analysis has been widely used to predict customer churn behavior. In this study, a customer churn analysis was conducted using publicly shared Telco telecommunications data. Predictive models were constructed using machine learning (LR, KNN, SVM, DT, RF, ANN), ensemble learning (XGBoost, Majority Voting), and deep learning (LSTM) methods. In addition, a 3-layered LSTM model was proposed. Accuracy (Acc), F1-score (F1), Precision (Prec), and Recall (Rec) rates were used to evaluate the models. As a result, the novel 3-layered LSTM model achieved 91.90% Acc, 91.49% Prec, 92.31% Rec, and 91.90% F1 values. The proposed model is competitive with the existing models. | |
| dc.identifier.citation | Başarslan, M. S., Ünal, A. & Kayaalp, F. (2025). A novel and robust LSTM model for customer churn analysis using deep, machine learning, and ensemble learning: A telecommunications case. Acta Infologica, 9(1), 55-73. https://doi.org/10.26650/acin.1584030 | |
| dc.identifier.doi | 10.26650/acin.1584030 | |
| dc.identifier.endpage | 73 | |
| dc.identifier.issn | 2602-3563 | |
| dc.identifier.issue | 1 | |
| dc.identifier.startpage | 55 | |
| dc.identifier.uri | https://hdl.handle.net/11363/10250 | |
| dc.identifier.volume | 9 | |
| dc.identifier.wos | 001433919400001 | |
| dc.identifier.wosquality | Q4 | |
| dc.indekslendigikaynak | Web of Science | |
| dc.institutionauthor | Ünal, Aslıhan | |
| dc.institutionauthorid | https://orcid.org/0000-0001-5896-8880 | |
| dc.language.iso | en | |
| dc.publisher | ISTANBUL UNIV, Rektorlugu, Beyazit, Fatih, ISTANBUL 34452, Turkiye | |
| dc.relation.ispartof | ACTA INFOLOGICA | |
| dc.relation.publicationcategory | Makale - Ulusal Hakemli Dergi - Kurum Öğretim Elemanı | |
| dc.rights | info:eu-repo/semantics/openAccess | |
| dc.subject | Customer Churn Analysis | |
| dc.subject | Ensemble Learning | |
| dc.subject | Machine Learning | |
| dc.subject | Deep Learning | |
| dc.subject | TECHNOLOGY::Information technology::Telecommunication | |
| dc.title | A Novel and Robust LSTM Model for Customer Churn Analysis Using Deep, Machine Learning, and Ensemble Learning: A Telecommunications Case | |
| dc.type | Article |
