https://github.com/miguellopezvirues/telecom_churn

Development and deployment of ML solution for a Telecom Churn business case.
https://github.com/miguellopezvirues/telecom_churn

aws churn-prediction deployment docker end-to-end machine-learning python

Last synced: 3 months ago
JSON representation

Development and deployment of ML solution for a Telecom Churn business case.

• Host: GitHub
• URL: https://github.com/miguellopezvirues/telecom_churn
• Owner: MiguelLopezVirues
• Created: 2024-09-07T21:49:51.000Z (almost 2 years ago)
• Default Branch: main
• Last Pushed: 2024-09-08T19:49:56.000Z (almost 2 years ago)
• Last Synced: 2025-02-10T00:19:34.183Z (over 1 year ago)
• Topics: aws, churn-prediction, deployment, docker, end-to-end, machine-learning, python
• Language: Jupyter Notebook
• Homepage:
• Size: 23.2 MB
• Stars: 0
• Watchers: 1
• Forks: 0
• Open Issues: 0
• Metadata Files:
  • Readme: README.md

Awesome Lists containing this project

README

          
# End to end ML solution for Telecom Churn prediction

This end to end project consists on providing a solution for a churn business case given a telecom churn dataset. Value is provided through:

- Data analysis: Understanding the root of the problem and providing actionable insights.

- ML model evaluation and configuration: Proposing several ML models and evaluating them, then proposing on the configuration of batch and real-time inference.

- Deployment of the ML solution: Containerizing the selected ML model to be deployed as a real-time inference solution in AWS Elasticbeanstalk.

## Install

To run the projects file, download it and run `pip install -r requirements.txt` for the home directory. Otherwise, ``Pipfile`` and ``Pipfile.lock`` are available inside `\Deployment`

## Using the application

To test the project, run `python predict_request-test.py`. It will make a request to the AWS application and return the churn prediction {0,1} for the example written on the script.

## Contents

- ``Churn_Business_Case.md``: Summary description of the business case for telecom churn.

- ``requirements.txt``

- ``\Development``

    - ``churn_all.csv``: Source dataset

    - ``Churn_ML_model_evaluation.ipynb``: Notebook with the exploratory data analysis and insights, evaluation and selection of the ML model, proposal of inference configuration and final recommendations.

- ``\Deployment``

    - ``train.py``: Data import from the source dataset, data processing and training of the selected model configuration from the Churn_ML_model_evaluation notebook.

    - ``predict_request.py``: Model loading and prediction, served as a flask application. Receives a JSON file with the variables for a single customer. Returns either 1 or 0 as a churn prediction.

    - ``predict_request-test.py``: Test of the request with the information of a single customer.

    - ``model.pkl``

    - ``category_group_map.pkl``: Category mapping automatically engineered through training, called by predict_request to preprocess the customer data.

    - ``Pipfile and Pipfile.lock``

    - ``Dockerfile``

## Technologies used

Python, Jupyter notebook, Pipenv, Docker, AWS Elasticbeanstalk