Nowadays, organizations are facing severalchallenges resulting from competition and markettrends. Customer churn is a real issue fororganizations in various industries, especially in thetelecommunications sector with a churn rate ofapproximately 30%, placing this industry in the top ofthe list. Because higher expenses are involved whentrying to attract a new customer than trying to retainan existing one, this is an important problem thatneeds an accurate resolution. This paper presents anadvanced methodology for predicting customers churnin mobile telecommunication industry by applying datamining techniques on a dataset consisting of call detailrecords. The data mining algorithms considered andcompared in this paper are Classification andRegression Tree, Chi-squared Automatic InteractionDetection Tree, and Quick Unbiased EfficientStatistical Tree.
展开▼
