This dataset contains a wealth of customer information collected from within a consumer credit card portfolio, with the aim of helping analysts predict customer attrition. It includes comprehensive demographic details such as age, gender, marital status and income category, as well as insight into each customer’s relationship with the credit card provider such as the card type, number of months on book and inactive periods. You can check the dataset here.
We use RandomForestClassifier to predict customer attrition. We also use several techniques such as GridSearchCV to tune the hyper-parameters of the RandomForestClassfier.