期刊名称：Annals of Dunărea de Jos University. Fascicle I : Economics and Applied Informatics
印刷版ISSN：1584-0409
出版年度：2018
期号：3
页码：1-7
出版社：Dunarea de Jos University of Galati
摘要：At the core of business lies customer satisfaction. However, customer retention strategies
are often based on individual preferences and conventional protocols. For an advantage in
the era of global competition, businesses require state-of-the-art techniques based on
information science and machine learning to correctly analyze historical data for the
prevention of customer loss. The present paper uses Rough Set theory to analyze customer
churn data for a telecom service provider. While this dataset has been analyzed in previous
research, this paper adds to the literature by taking a systematic and comprehensive
approach to the selection of significant features, using them to infer a set of rules clearly
describing customer groups that are most likely to churn, and drawing appropriate
conclusions from the rules.
