Customer churn analysis with data mining techniques in direct sales sector

All representative data is obtained from a direct sales company for this study. Customer loss has been calculated, by classification algorithms which are data mining methods, with the data provided. Cross Industry Standard Process Model (CRISP) steps is followed for the data mining while analyzing. Decision trees have been chosen from classification algorithms. In addition to this C4.5 decision tree and Gini decision tree algorithms were used in this study. The performance of the models obtained by C4.5 decision tree and Gini decision tree algorithms were measured and evaluated by hold-out performance method. The data set is separated with hold-out method by %90-%10, %80-%20, %70-%30, %60-%40 respectively. R programming language has been used for the analysis conducted.