A churn prediction model for prepaid customers in telecom using fuzzy classifiers

Abstract

The incredible growth of telecom data and fierce competition among telecommunication operators for customer retention demand continues improvements, both strategically and analytically, in the current customer relationship management (CRM) systems. One of the key objectives of a typical CRM system is to classify and predict a group of potential churners form a large set of customers to devise profitable and targeted retention campaigns for keeping a long-term relationship with valued customers. For achieving the aforementioned objective, several churn prediction models have been proposed in the past for the accurate identification of the customers who are prone to churn. However, these previously proposed models suffer from a number of limitations which place strong barriers towards the direct applicability of such models for accurate prediction. Firstly, the feature selection methods adopted in majority of the past work neglected the information rich variables present in call details record for model development. Secondly, selection of important features was done through statistical methods only. Although statistical methods have been applied successfully in diverse domains, however, these methods alone without the augmentation of domain knowledge have the tendency to yield erroneous results. Thirdly, the previous models have been validated mainly with benchmark datasets which do not provide a true representation of real world telecom data con-B Muhammad Usman

Enhanced Churn Prediction in theTelecommunication Industry

“…More so, the result showed that internet service, types of contract entered, internet security were major factors that influence churn. The study used various methods of classifiers like earlier researchers [6] [7] [11] who used various methods (Artificial Neural Networks and Decision Trees). The findings from this study showed that the method used were all effective and can be equally strong to predict churn.…”

“…In terms of variables that causes churn the findings of this study agree with [15] in that many of the variables have correlations with churn and affects it, however, internet services and types of contract affect churn the more. In studies where classifications were not carried out like the study by [7] that adopted a working methodology of Ensemble based Classifiers such as bagging, boosting and random forest, in contrast analysis to common classifiers such as; Decision Tree, Naïve Bayes Classifier and Support Vector Machine. The study concluded that effectiveness is best with simple classifiers like SVM and logistic regression but the result from logistic regression showed that it was the best Classifier for the Churn Prediction Problem as compared to other models.…”

A Critical Examination of Different Models for Customer Churn Prediction using Data Mining

“…Development of the www service has led to the growth of e-commerce. E-Commerce uses ICT in customer service and business operations for creating, converting and redefining relationships between buyers, sellers [2]. This made it possible for businesses to become close to potential and existing consumers and to develop the more loyal relationship between them.…”

“…Data mining is used to extract data patterns such as the data record groups, unusual data records and data interdependencies addressed using cluster analysis, anomaly detection and association rule mining respectively. Data mining normally consist of six tasks [2] namely (i) anomaly detectionto identify unusual data records (ii) association rule miningto search for relationships between variables (iii) clusteringto discover structures in data (iv) classificationto generalize to apply to new data (v) regressionto estimate relationships among datasets and model the data with least error (vi) summarizationto represent the data in compact form.…”

Prediction of users charging time in cloud environment using machine learning

“…The classifiers used before are not the best methods while dealing with a noise data, and the proposed model shows a higher TP values compared to other models. A number of predominant classifiers namely, Neural Network, Linear regression, support vector machines, Gradient Boosting and Random Forest are compared with fuzzy classifiers to highlight the superiority of fuzzy classifiers in predicting the accurate set of churners in [10].…”