Churn-prediction
Churn-prediction
Data preprocessing:: Label encoder is used to normalize the categorical variable
Data Transformation:: For each data transformation sepeate fiels are used Base line (RAW) is used for Original dataset without any data transformation
Feature selection:: Univarient feature selection is used for feature selection
Paremeter tunning:: Grid Search CV is used for parameter tuning for the classifiers
Dataset:: Three datasets are used in our study
- data set 1 Telecom_customer churn (10000).rar
- churn-5000.csv
- churn-data-333.csv