Customer Churn Prediction with Machine Learning Methods In Telecommunication Industry

Year 2025, Volume: 5 Issue: 1, 32 - 41, 16.06.2025

Buse Demir , Övgü Öztürk Ergün

https://doi.org/10.54569/aair.1709274

Abstract

With the emergence of new competitors and increasing investments in telecommunication services, change often occurs and hence importance of marketing strategies and customer behavior prediction have become an important demand for companies. New regulations and technologies increase competition among mobile operators. Since acquiring a new customer is more expensive than acquiring active customers, companies seek solutions to reduce the churn rate. Therefore, telecommunications companies want to analyze the concept of the customer's desire to change service provider and take necessary measures to protect their existing customers. In this study, usage information, usage trends, subscription commitment, subscription age, ARPU and billing information, competitor familiarity, outgoing call information, number porting experience, etc. Loss estimation modeling is taken into account. Dataset includes 593 columns and 1826588 lines. Corporate mobile customers are analyzed by dividing into three subgroups as Single Line Mobile Customers, 2-5 Line Mobile Customers, and 6-15 Line Mobile Customers. In order to estimate customer loss, four different ML methods are used while creating loss prediction models. The model is developed by using 600 different variables and loss estimation. ROC curves and lift chart results for different corporate mobile customer groups are compared and the most suitable models are depicted.

Keywords

Churn Prediction , Machine Learning , Telecommunication , Customer , Data Analysis , Random Forests , AI

Telekomünikasyon Sektöründe Makine Öğrenme Yöntemleriyle Müşteri Kaybı Tahmini

Abstract

Yeni rakiplerin ortaya çıkması ve telekomünikasyon hizmetlerine yapılan yatırımların artmasıyla birlikte sıklıkla değişim yaşanmakta ve dolayısıyla pazarlama stratejilerinin ve müşteri davranış tahminlerinin önemi şirketler için önemli bir talep haline gelmiştir. Yeni düzenlemeler ve teknolojiler mobil operatörler arasındaki rekabeti artırmaktadır. Yeni bir müşteri edinmek, aktif müşteri edinmekten daha maliyetli olduğundan şirketler, müşteri kaybını azaltmak için çözümler aramaktadır. Bu nedenle, telekomünikasyon şirketleri müşterinin servis sağlayıcısını değiştirme isteği kavramını analiz etmek ve mevcut müşterilerini korumak için gerekli önlemleri almak istemektedir. Bu çalışmada, kullanım bilgileri, kullanım eğilimleri, abonelik taahhüdü, abonelik yaşı, ARPU ve faturalama bilgileri, rakip aşinalığı, giden arama bilgileri, numara taşıma deneyimi vb. kullanılarak kayıp tahmin modellemesi yapılmıştır. Veri seti 593 sütun ve 1826588 satır içermektedir. Kurumsal mobil müşteriler, Tek Hatlı Mobil Müşteriler, 2-5 Hatlı Mobil Müşteriler ve 6-15 Hatlı Mobil Müşteriler olmak üzere üç alt gruba ayrılarak analiz edilmiştir. Müşteri kaybını tahmin etmek için, kayıp tahmin modelleri oluşturulurken dört farklı ML yöntemi kullanılmıştır. Model 600 farklı değişken ve kayıp tahmini kullanılarak geliştirilmiştir. Farklı kurumsal mobil müşteri grupları için ROC eğrileri ve kaldırma grafiği sonuçları karşılaştırılarak en uygun modeller gösterilmiştir.

Keywords

Kayıp Tahmini , Makine Öğrenmesi , Telekomünikasyon , Rastgele Orman , Veri Analizi , YZ , Müşteri

References

• Gumus N. "Determination of Smartphone Users’ Perceptions of Branded Mobile Applications in Turkey", Online Journal of Communication and Media Technologies, (2017) 27-45.
• Kumar V, Petersen JA. "Statistical Methods in Customer Relationship Management", Wiley (2012).
• Keaveney, S., “Customer switching behavior in service industries”, Journal of Marketing, (1995) 59 (2), 71-82.
• Rosenberg L et al. “A marketing approach for customer retention”, J of Consumer Marketing (1992) 1(2): 45-51.
• Grönroos C. "Relationship Marketing: The Strategy Continuum”, Journal of the Academy of Marketing Science, (1995) 23 (4), 252-25.
• Akan O, Verma A. “Machine Learning Models for Customer Churn Risk Prediction”, IEEE 13th Annual Ubiquitous Computing, Electronics & Mobile Communication Conference (2022) 0623-0628.
• Ghosh, A., Dehuri, S., Ghosh, S. (eds.): “Multi-objective evolutionary algorithms for knowledge discovery from databases”, Springer, Berlin (2008).
• Lee J, et al. "The impact of switching costs on the customer satisfaction‐loyalty link: mobile phone service in France", Journal of Services Marketing, (2001) 35-48.
• Hung S, Yen DC, Wang H. “Applying Data Mining to Telecom Churn Management”, Expert Systems with Applications, (2006) 31, 515-524.
• Gupta S, et al. “Valuing Customers”, Journal of Marketing Research, (2004) 7-18.
• Poel V, et al. "Customer attrition analysis for financial services using proportional hazard models," European Journal of Operational Research, (2004) 157: 196-217.
• Kamalraj, Malathi, “A Survey on Churn Prediction Techniques in Communication Sector”, International Journal of Computer Applications, (2013) 64 (5), 39-41.
• Ahmad AK, LJafar A, Aljoumaa K. “Customer churn prediction in telecom using machine learning in big data platform”, Journal of Big Data (2019) 6:(28).
• Chouriek A, El Haj EHI. ”Deep Convolutional Neural Networks for Customer Churn Prediction Analysis”, International Journal of Cognitive Informatics and Natural Intelligence (2020) 14(1):1-16.
• Kilimci ZH, et al., “Sentiment Analysis Based Churn Prediction in Mobile Games using Word Embedding Models and Deep Learning Algorithms”, 2020 International Conference on Innovations in Intelligent SysTems and Applications (INISTA), (2020) 1-7.
• Nazeer WW, et al. “Predictive analytics using big data for increased customer loyalty: Syriatel Telecom Company case study”, Journal of Big Data, (2020) 7:29.
• B. Huang, M. T. Kechadi, and B. Buckley,”Customer churn prediction in telecommunications.” Expert Systems with Applications (2012)1414-1425.
• Loh, WY, “Classification and regression trees”, WIREs Data Mining and Knowledge Discovery (2011).
• Liu, Y. "Design and evaluation of visualization support to facilitate decision trees classification", International Journal of Human - Computer Studies (2007) 95-110.
• Yang C, et al. "Estimation of Lyapunov exponents from a time series for n-dimensional state space using nonlinear mapping", Nonlinear Dynamics (2012).
• Li H, et al. “Supervised Massive Data Analysis for Telecommunication Customer Churn Prediction”. BDCloud-SocialCom-SustainCom 2016: 163-169.
• Sonmez B, Sarikaya AA, Bahtiyar S. "Machine Learning based Side Channel Selection for Time-Driven Cache Attacks on AES", 4th International Conference on Computer Science and Engineering (2019) 1-5.
• Chang Q, et al. “An overview of artificial intelligence: basics and state-of-the-art algorithms" Elsevier BV (2021)
• Tin Kam Ho,Random “Decision Forests” (PDF). Proceedings of the 3rd International Conference on Document Analysis and Recognition, Montreal, (1995).
• Bilen A, Özer AB. "Cyber- attack method and perpetrator prediction using machine learning algorithms", PeerJ Computer Science, (2021).
• Yabas U, Cankaya HC, “Churn prediction in subscriber management for mobile and wireless communications services”, 2013 IEEE Globecom Workshops, (2013) 991-995.
• Demir Buse, “Churn prediction of enterprise mobile customers with ML methods in telecommunication industry”, YÖK Ulusal Tez Merkezi (2021), No: 685065.