Customer Churn Prediction model created based on Classification algorithms created with reference to DataTalks.Club mlzoomcamp course created using Google Colab
Reading the data from csv file and checking for null values and examining important features.
- Examining important features using mutual information(MI) scores.
- Changing categorial variables to numerical variables using one hot encoding and pandas dummy
- Checking correlation with numerical variables
Models |
---|
Logistic Regression |
Random Forest classifier |
XGB classifier |
Picking the best model based on maximum Accuracy
Predicting the customer churn using the final model with test data