In this repository I explore different methods to find out which features are most important in the customer churn task.
There are 3 different methods, that I chose:
- using Logistic Regression
- using CatBoost Classifier
- and using Ensemble of different models (Logistic Regression, HistGradBoostTree and Naive Bayes model)
- PCA (Principal Component Analysis) method
To make this project, I was using this cite. It was very helpfull for me.