🧠 Stroke Risk Prediction

This project was developed as my final project during my internship at Aman Holding.

It is a Streamlit web application that predicts whether a person is at risk of stroke based on healthcare data, while also providing interactive visualizations, dataset insights, and model explainability using SHAP values.

---

🎥 Demo

---

📊 Project Overview

• Uses a healthcare dataset from Kaggle for training and evaluation.

• Provides interactive visualizations to explore dataset patterns and insights.

• Integrates a trained Machine Learning model to predict stroke likelihood.

• Uses SHAP values to explain the model's predictions.

• Includes a Jupyter Notebook documenting:

  • Data preprocessing
  • Feature engineering
  • Model selection
  • Model evaluation

---

🚀 Features

• Stroke Prediction: Enter patient health attributes and receive a prediction.
• Visual Analytics: Explore dataset statistics and visualizations.
• Explainable AI: SHAP values help explain why the model made a prediction.
• End-to-End Pipeline: From raw data → preprocessing → model → deployment.

---

🛠️ Tech Stack

• Python 3.8+
• NumPy
• Pandas
• Scikit-learn
• SHAP
• Matplotlib
• Seaborn
• Plotly
• Streamlit
• Joblib
• PIL

---

📦 Installation

Clone the repository:

git clone https://github.com/Seif-Elmezaien/Stroke-Detection.git
cd Stroke-Detection

Install dependencies

pip install -r requirements.txt

---

▶️ Usage

Run the Streamlit application:

streamlit run app.py

The application will open in your browser.

---

📁 Project Structure

Stroke-Detection/
├── demo/
│   └── stroke_demo.gif
├── Notebook/
├── app.py
├── requirements.txt
└── README.md

---

⚠️ Disclaimer

This project was developed for educational purposes as part of my internship at Aman Holding. The dataset is publicly available on Kaggle.

The application is not intended for medical diagnosis or to replace professional medical advice.