期刊名称：International Journal of Innovative Research in Computer and Communication Engineering
印刷版ISSN：2320-9798
电子版ISSN：2320-9801
出版年度：2017
卷号：5
期号：3
页码：6157
DOI：10.15680/IJIRCCE.2017.0503381
出版社：S&S Publications
摘要：The interest for big data mining techniques has increased tremendously in the recent researches.Numerous classification and clustering techniques based on both supervised and unsupervised learning models wereproposed. And these models had been applied in a wide range of business applications related to customer managementto determine the optimal cluster for dynamic data. However, when dealing with big data in the industry, existing churnprediction models cannot work very well. In addition, decision makers are always faced with imprecise operationmanagement. Hence the need for prediction mechanism for various strategies has become more important in large scaledata source and streams. Hence, Big Data clustering algorithm called semantic-driven subtractive clustering method(SDSCM) [1] which is a combination of Axiomatic Fuzzy Sets (AFS) and Subtractive clustering method (SCM) is usedto classify the customer churn rate in Telecom industry. In this work, we compareSDSCM with parallel k-mean andparallel k-median classifiers to cluster the records of the customer information or log details for maximizing thecustomer retention in the Telco service. The experimental result aims to predict which combination of algorithm worksgood based on accuracy of classifying customer in terms of loyalty.
