Powered by OpenAIRE graph
Found an issue? Give us feedback
image/svg+xml art designer at PLoS, modified by Wikipedia users Nina, Beao, JakobVoss, and AnonMoos Open Access logo, converted into svg, designed by PLoS. This version with transparent background. http://commons.wikimedia.org/wiki/File:Open_Access_logo_PLoS_white.svg art designer at PLoS, modified by Wikipedia users Nina, Beao, JakobVoss, and AnonMoos http://www.plos.org/ Istanbul Ticaret Uni...arrow_drop_down
image/svg+xml art designer at PLoS, modified by Wikipedia users Nina, Beao, JakobVoss, and AnonMoos Open Access logo, converted into svg, designed by PLoS. This version with transparent background. http://commons.wikimedia.org/wiki/File:Open_Access_logo_PLoS_white.svg art designer at PLoS, modified by Wikipedia users Nina, Beao, JakobVoss, and AnonMoos http://www.plos.org/
addClaim

Churn prediction in telecommunication sector

Authors: Hassan Mohamed, Faiza;

Churn prediction in telecommunication sector

Abstract

ÖZET Literatürde, Makine Öğrenimi, Veri Madenciliği ve Hibrit teknikleri gibi farklı teknikler kullanılarak Ayrılma/Çalkalanma Tahmini gerçekleştirilmiştir. Bu teknikler, şirketleri ve işletmeleri, hizmetlerini kullanarak şirketlerinde kalabilmeleri için müşterileri belirleme ve tahmin etme ayrıca ayrılan müşteri konusunda destekler. Üst düzey yöneticilerin ve karar vericilerin güvenilir kararlar almasına ve Müşteri İlişkileri Yönetimi (CRM) departmanına da yardımcı olur. Bu çalışmada, müşteri kaybını tahmin etmek için Orange (Uluslararası Telekominikasyon firması) adlı bir telekom sektörü ayrılan müşteri veri seti kullanılmıştır. Topluluk sınıflandırıcıları AdaBoostM1, PCA, InfoGain, Gain Ratio, Bagging ile birlikte J4. 8, Naive Bayes, Lojistik Regresyon, Rastgele Orman, KNN, LMT (Lojistik model Ağacı) sınıflandırıcıları kombinasyonları ile birlikte kullanılır. Torbalama ve J4. 8 kombinasyonu ile en yüksek % 94 doğruluk elde edilir. Sonuçlar diğer çalışmalarla da karşılaştırılmış ve bu çalışma araştırılan literatür kadar iyi performans göstermiş ve bazı vakalarda daha başarılı olduğu görülmüştür. Anahtar Kelimeler: Ayrılama Tahmini, kNN, LMT, CRM, PCA ABSTRACT Churn Prediction has been implemented in the research works and more studies on it been published using different advanced mechanisms including Machine Learning, Data Mining, and Hybrid mechanism. These mechanisms found out to help enterprise companies and small businesses to classify and predict churning customers to be able to retain them to stay with their company using their services. Also, found out to help top managers and decision makers to be able take reliable decisions and Customer Relation Management CRM department as well. In this study, a telecom sector churn dataset named Orange which belongs to International Orange Telecom Company is used for customer churn prediction. Ensemble classifiers are used AdaBoostM1, PCA, Gain Ratio, Info Gain, Bagging in combination with J4. 8, Naïve Bayes, Logistic Regression, Random Forest, KNN, LMT (Logistic model Tree). Highest accuracy of 94% is obtained by combination of Bagging and J48. The results are compared with other studies as well and this study performed as good as the surveyed literature and surpassed in same cases. Keywords: Churn Prediction, KNN, LMT, CRM, PCA İÇİNDEKİLER TABLE OF CONTENTS . ii ABSTRACT . iii ÖZET . iii ACKNOWLEDGEMENT . iiv LIST OF FIGURES . v LIST OF TABLES . vi ABBREVIATIONS . vii 1. INTRODUCTION . 1 1. 1 Background of the Study . 1 1. 2. Problem Statement . 4 2. LITERATURE REVIEW . 5 3. MATERIAL AND METHODOLOGIES . 12 3. 1 Data Description . 12 3. 1. 1 Brief history of orange telecom company . 12 3. 1. 2 Orange telecom dataset . 13 3. 2 Methodologies . 14 3. 2. 1 Naïve bayes . 14 3. 2. 2 J48 . 15 3. 2. 3 Random forests . 16 3. 2. 4 K Nearest Neighbor (kNN) . 16 3. 2. 5 Logistics regression . 17 3. 2. 6 Decision stump . 17 3. 2. 7 Logistic model tree LMT . 17 3. 2. 8 AdaboostM1 . 18 3. 2. 9 Principal component analysis (PCA) . 18 3. 2. 10 Gain ratio . 18 3. 2. 11 Information gain . 18 3. 2. 12 Bagging . 18 3. 2. 13 WEKA . 18 4. EVALUATIONS AND DISCUSSION . 21 4. 1 Implementation Results Evaluation . 21 4. 2 Discussion . 26 5. CONCLUSION . 27 REFERENCES . 29 BIOGRAPHY . 32

Tez (Yüksek Lisans) -- İstanbul Ticaret Üniversitesi -- Kaynakça var.

Country
Turkey
Related Organizations
Keywords

Prediction theory, Tahmin teorisi, Karar verme, Decision making, T 57.95/H37

  • BIP!
    Impact byBIP!
    selected citations
    These citations are derived from selected sources.
    This is an alternative to the "Influence" indicator, which also reflects the overall/total impact of an article in the research community at large, based on the underlying citation network (diachronically).
    0
    popularity
    This indicator reflects the "current" impact/attention (the "hype") of an article in the research community at large, based on the underlying citation network.
    Average
    influence
    This indicator reflects the overall/total impact of an article in the research community at large, based on the underlying citation network (diachronically).
    Average
    impulse
    This indicator reflects the initial momentum of an article directly after its publication, based on the underlying citation network.
    Average
Powered by OpenAIRE graph
Found an issue? Give us feedback
selected citations
These citations are derived from selected sources.
This is an alternative to the "Influence" indicator, which also reflects the overall/total impact of an article in the research community at large, based on the underlying citation network (diachronically).
BIP!Citations provided by BIP!
popularity
This indicator reflects the "current" impact/attention (the "hype") of an article in the research community at large, based on the underlying citation network.
BIP!Popularity provided by BIP!
influence
This indicator reflects the overall/total impact of an article in the research community at large, based on the underlying citation network (diachronically).
BIP!Influence provided by BIP!
impulse
This indicator reflects the initial momentum of an article directly after its publication, based on the underlying citation network.
BIP!Impulse provided by BIP!
0
Average
Average
Average
Green