🧠 Brain Tumor Classification

This repository contains a deep learning-powered web application for classifying brain tumors using MRI images. The system uses Transfer Learning with VGG16, modular ML pipelines, and Flask for deployment. Project pipelines are managed and version-controlled using DVC (Data Version Control).

🚀 Features

• VGG16 Transfer Learning for brain tumor classification
• Modular architecture: ingestion, training, evaluation, prediction
• Flask web interface for real-time image classification
• DVC-integrated data and model versioning
• Conda-based environment setup

---

📁 Project Structure

├── artifacts/ # Stored artifacts (e.g., trained models)
├── config/ # Configuration files
├── logs/ # Logging info
├── research/ # Notebooks and experiments
├── src/ # Core package source code
├── templates/ # HTML templates for Flask app
├── .dvc/ # DVC configuration
├── app.py # Flask application
├── main.py # Pipeline execution script
├── params.yaml # Training/evaluation parameters
├── dvc.yaml # DVC pipeline file
├── requirement.txt # Required Python packages
├── README.md # Project documentation
└── setup.py # Package setup

🧪 Setup Instructions

✅ Create Conda Environment

conda create -n tumor-classifier python=3.10 -y
conda activate tumor-classifier

📦 Install Requirements

pip install -r requirement.txt

💾 DVC Setup (Optional but recommended)

dvc init
dvc repro

💻 Run the Flask Web App

python app.py

---

🖼️ Preview

🙋‍♀️ Author

Mamoona Ramzan (Software Engineering Student at NUST)