Skip to content

KhushiSindhu/Telecome-Churn-PCA-And-Multiple-Models

Repository files navigation

Project Name

Telecome-Churn Prdeiction

Table of Contents

General Information

  • Background:

  • Business Problem:

  • DataSet :

Conclusions

We build various models, using Ridge and Lasso Reguralization . And choose the model with Lasso regularisation with optimal value of alpha as final model.

The below mentioned variables are Top 10 significant in predicting the churn

Postive Features

Negative Features

Acknowledgements

  • This project is based on upgrad Course PCA, Tree Models (Random Forest, ADABoost, GD Boost, XGBoost)

Technologies Used

  • Python - version 3.9.7
  • Jupyter Notebook Server - version 6.4.5

Contact

Created by @KhushiSindhu - feel free to contact me!