Churn Prediction Project

State of deployment

The lunch time using the AWS EC2 instance was from: 2024/11/07 16:57 GMT-5 to 2024/11/15 GTM-5 8:48. Due to another current project being develop the app is not running

Overview

The current project is deployed using AWS and docker. This project provides a machine-learning model for predicting customer churn. It includes the necessary scripts and configurations to deploy the model as a web application, suitable for Docker-based or local deployment. The project leverages Python with dependencies managed through Poetry.

Directory Structure

• app.py: Main application file to run the churn prediction model or API.
• Dockerfile: Contains instructions for setting up a Docker container to run the application.
• requirements.txt: Lists the Python dependencies.
• utils.py: Contains utility functions supporting the main application.

Prerequisites

• Python: Ensure Python is installed (recommended: Python 3.8 or above).
• Poetry: Used for dependency management. Install with pip install poetry.
• Docker: Optional, for containerized deployment.

Installation

1. Clone the repository.

2. Install dependencies using Poetry:

   poetry install

   pip install -r requirements.txt

manual installation with requirements.txt:

pip install -r requirements.txt

Running the Application

To start the application:

python app.py

For Docker users:

Build the Docker image:

docker build -t churn-prediction .

Run the Docker container:

docker run -p 8000:8000 churn-prediction

Usage

Once the application is running, access it locally or through the specified port if using Docker.

Additional Files

• .gitignore: Lists files to exclude from version control.
• .replit: Contains configuration for running the app on Replit.
• poetry.lock, pyproject.toml: Configuration files for Poetry.