🍷End-to-End Machine Learning Project: Wine Quality Prediction

Welcome to the Wine Quality Prediction project! This repository contains an end-to-end machine learning pipeline built using the Wine Quality Dataset. The project leverages tools like MLflow, DagsHub, and Flask to create a robust and interactive solution for predicting wine quality.

---

Table of Contents

• Project Overview
• Features
• Tech Stack
• Project Workflow
• Folder Structure
• Setup Instructions
• Usage
• License

---

Project Overview

This project demonstrates the complete lifecycle of a machine learning solution:

1. Data Ingestion: Collect and preprocess the Wine Quality dataset.
2. Model Training: Train machine learning models to predict wine quality.
3. Model Evaluation: Evaluate the model's performance using metrics and track experiments with MLflow and DagsHub.
4. Deployment: Deploy the model using Flask to provide a simple and interactive user interface.

---

Features

• End-to-End ML Pipeline: From data ingestion to deployment.
• Experiment Tracking: Integrated with MLflow and DagsHub for tracking experiments and model performance.
• Interactive UI: A simple Flask-based web interface for users to input data and get predictions.
• Configurable Workflow: YAML-based configuration for easy customization.

---

Tech Stack

• Programming Language: Python
• Libraries: Pandas, NumPy, Scikit-learn, Flask
• Experiment Tracking: MLflow, DagsHub
• Deployment: Flask
• Version Control: Git and GitHub
• Containerization: Docker

---

Project Workflow

The project is divided into the following stages:

1. Data Ingestion: Load and preprocess the dataset.
2. Data Validation: Validate the dataset schema using schema.yaml.
3. Data Transformation: Perform feature engineering and preprocessing.
4. Model Training: Train and save the model.
5. Model Evaluation: Evaluate the model and log metrics using MLflow.
6. Deployment: Deploy the model using Flask for real-time predictions.

Steps to Run the Workflow

1. Update the following configuration files:
   • config.yaml
   • schema.yaml
   • params.yaml
2. Implement the pipeline components in the src directory.
3. Run the pipeline using main.py.

---

Folder Structure

DsProject/
├── .github/workflows/ # CI/CD workflows
├── config/ # Configuration files (config.yaml, schema.yaml, params.yaml)
├── research/ # Notebooks for exploratory data analysis
├── src/datascience/ # Source code for ML pipeline
├── templates/ # HTML templates for Flask UI
├── app.py # Flask application
├── main.py # Entry point for the ML pipeline
├── requirements.txt # Python dependencies
├── Dockerfile # Docker configuration
├── setup.py # Package setup
└── README.md # Project documentation

Setup Instructions

1. Clone the repository:

   git clone https://github.com/adiManethia/DsProject.git
   cd DsProject
   

2. Create a virtual environment and activate it:

   python -m venv venv
   source venv/bin/activate  # On Windows: venv\Scripts\activate
   

3. Install dependencies:

   pip install -r requirements.txt
   

4. Run the pipeline:

   python main.py
   

5. Start the Flask app:

   python app.py
   

Usage

• Experiment Tracking : Use MLflow and DagsHub to monitor model performance.
• Web Interface : Input wine features through the Flask UI to get quality predictions.

License

This project is licensed under the GPL-3.0 License.

Feel free to contribute to this project by submitting issues or pull requests.