Customer churn prediction system: a machine learning approach

Most popular predictive models have been applied, namely, logistic regression, naive bayes, support vector machine, random forest, decision trees, etc. on train set as well as boosting and ensemble techniques are applied to see the effect on accuracy of models.