Customer Churn Use Case

In this use case repository, there is the preprocessing and the model construction for a binary classification problem of whether a customer will churn or not.

Data

We have received two main datasets from MYT, one the incoice oriented and the other about customer complaints. This data has been processed, cleaned and transformed.

Use of the pipeline

The pipeline is used in Mage divided in blocks like the energy prediction model

Notebooks

• model.ipynb: This notebook has the model construction
• model.ipynb: This notebook has the preprocessing

The model used for this use case is the churn_model and it is inside the folder models

How to run

Install the project's dependencies from 'requirements.txt'