Comparing to Techniques Used in Customer Churn Analysis

Main Article Content

Özer Çelik Usame Ömer OSMANOĞLU


In today's competitive conditions, the importance of minimizing costs is increasing day by day. As a result of the researches, it has been determined that the cost of attracting new customers is 10 times more than the cost of holding existing customers. This increases the importance of customer churn analysis, too. In this review, machine learning algorithms such as artificial neural networks (ANN), decision trees, support vector machines (SVM), naive bayes, k-nn and extreme gradient boosting (XGBoost) and customer churn analysis, Cox proportional hazard model and deep learning techniques. In addition, customer churn analysis studies conducted in various sectors using these techniques were examined. When the customer churn analysis studies are examined, it is seen that more complex systems can get modeling and reached higher success rates with deep learning technique. However, because it is a new technique and could not give stable results, machine learning algorithms, which are the closest alternative, are thought to be able to useful in estimating time-related events such as customer churn. The Cox regression model was found to be successful in estimating the independent variables affecting the time variable, the rate of life expectancy and the groups under risk. It is seen that deep learning techniques give better results in more complex structures.

Article Details

How to Cite
ÇELIK, Özer; OSMANOĞLU, Usame Ömer. Comparing to Techniques Used in Customer Churn Analysis. Journal of Multidisciplinary Developments, [S.l.], v. 4, n. 1, p. 30-38, july 2019. ISSN 2564-6095. Available at: <>. Date accessed: 15 july 2024.
Natural Sciences - Regular Research Paper


[1] Nettleton, D. (2014). Commercial data mining: processing, analysis and modeling for predictive analytics projects. Elsevier.

[2] Kotler, P., & Keller, K. L. (2009). Manajemen pemasaran.

[3] Seker, S. E. (2016). Müşteri Kayıp Analizi (Customer Churn Analysis). YBS Ansiklopedi, 3(1), 26-29.

[4] Poel, V. D., Lariviére (2004). Customer Attrition Analysis For Financial Services Using Proportional Hazard Models. European Journal of Operational Research 157: 196–217. doi:10.1016/s0377-2217(03)00069-9. CiteSeerX:

[5] Erdem, E. S. (2014). Ses sinyallerinde duygu tanıma ve geri erişimi (Master's thesis, Başkent Üniversitesi Fen Bilimleri Enstitüsü).

[6] Topal, C. (2017). Alan Turing'in Toplumbilimsel Düşünü: Toplumsal Bir Düş Olarak Yapay Zekâ. DTCF Dergisi, 57(2).

[7] Cox, D. R., & Oakes, D. (1984). Analysis of survival data. Chapman&Hall, London.

[8] Kaplan, E. L., & Meier, P. (1958). Nonparametric estimation from incomplete observations. Journal of the American statistical association, 53(282), 457-481.

[9] Mantel, N. (1966). Evaluation of survival data and two new rank order statistics arising in its consideration. Cancer Chemother Rep, 50, 163-170.

[10] Cox, D. R. (1992). Regression models and life-tables. In Breakthroughs in statistics (pp. 527-541). Springer, New York, NY.

[11] Fleming, T. R., & Lin, D. Y. (2000). Survival analysis in clinical trials: past developments and future directions. Biometrics, 56(4), 971-983.

[12] Schmidhuber, J. (2015). Deep learning in neural networks: An overview. Neural networks, 61, 85-117.

[13] Huigevoort, C., & Dijkman, R. (2015). Customer churn prediction for an insurance company.

[14] Sirmacek, B. (2007). FPGA ile mobil robot için öğrenme algoritması modellenmesi (Doctoral dissertation).

[15] Altunisik, R. (2015). Büyük Veri: Fırsatlar Kaynağı mı Yoksa Yeni Sorunlar Yumağı mı?. Yildiz Social Science Review, 1(1).

[16] Gor, İ. (2014). Vektör nicemleme için geometrik bir öğrenme algoritmasının tasarımı ve uygulaması (Master's thesis, Adnan Menderes Üniversitesi).

[17] Turkmenoglu, C. (2016). Türkçe Metinlerde Duygu Analizi (Doctoral dissertation, Fen Bilimleri Enstitüsü).

[18] Tamaddoni, A., Stakhovych, S., & Ewing, M. (2016). Comparing churn prediction techniques and assessing their performance: a contingent perspective. Journal of service research, 19(2), 123-141.

[19] Kocadayi, Y., Erkaymaz, O., & Uzun, R. (2017). Yapay Sinir Ağları ile Tr81 Bölgesi Yıllık Elektrik Enerjisi Tüketiminin Tahmini. BİLDİRİ ÖZETLERİ KİTABI, 239.

[20] Albayrak, A. S., & Yilmaz, Ö. G. Ş. K. (2009). Veri madenciliği: Karar ağacı algoritmaları ve İMKB verileri üzerine bir uygulama. Süleyman Demirel Üniversitesi İktisadi ve İdari Bilimler Fakültesi Dergisi, 14(1).

[21] Guneren, H. (2015). Destek vektör makineleri kullanarak gömülü sistem üzerinde yüz tanıma uygulaması (Doctoral dissertation).

[22] Ozkan, H. (2013). K-Means Kümeleme ve K-NN Sınıflandırma Algoritmalarının Öğrenci Notları Ve Hastalık Verilerine Uygulanması.

[23] Brownlee J. (2016). A Gentle Introduction to XGBoost for Applied Machine Learning. Recieved from machine-learning/ Accessed 04 January 2019

[24] Collett, D. (1994). Modelling survival data. In Modelling Survival Data in Medical Research (pp. 53-106). Springer US.

[25] Jamal, Z., & Bucklin, R. E. (2006). Improving the diagnosis and prediction of customer churn: A heterogeneous hazard modeling approach. Journal of Interactive Marketing, 20(3-4), 16-29.

[26] Wong, K. K. K. (2011). Using cox regression to model customer time to churn in the wireless telecommunications industry. Journal of Targeting, Measurement and Analysis for Marketing, 19(1), 37-43.

[27] Martinsson, E. G. I. L. (2016). Wtte-rnn: Weibull time to event recurrent neural network (Doctoral dissertation, Master’s thesis, University of Gothenburg, Sweden).

[28] Castanedo, F., Valverde, G., Zaratiegui, J., & Vazquez, A. (2014). Using deep learning to predict customer churn in a mobile telecommunication network.

[29] Wangperawong, A., Brun, C., Laudy, O., & Pavasuthipaisit, R. (2016). Churn analysis using deep convolutional neural networks and autoencoders. arXiv preprint arXiv:1604.05377.

[30] Spanoudes, P., & Nguyen, T. (2017). Deep learning in customer churn prediction: unsupervised feature learning on abstract company independent feature vectors. arXiv preprint arXiv:1703.03869.