Heart Stroke Risk Prediction System

An End-to-End Machine Learning Solution for Clinical Risk Assessment

Executive Summary

This project provides a data-driven approach to identifying high-risk stroke patients. By leveraging clinical parameters such as blood pressure, glucose levels, and smoking history, the system outputs a probability score using a trained Logistic Regression model.

Technical Stack

• Core Engine: Python 3.10+
• Data Science: Scikit-Learn (Model & Scaler), Pandas (EDA), NumPy
• Web Framework: Streamlit (UI/UX)
• Deployment: Streamlit Cloud / GitHub Actions

Data Engineering & Features

The model was trained on the [Kaggle Stroke Dataset]. Key technical implementations include:

• Feature Engineering: Created pulse_pressure (SysBP - DiaBP) and age_glucose_impact to capture non-linear risks.
• Preprocessing: Robust handling of class imbalance and StandardScaler normalization.
• Pipeline: Integrated a serialized joblib pipeline for seamless deployment.

Local Setup

# Clone the repository
git clone [https://github.com/YourUsername/Heart-disease-predictor.git](https://github.com/YourUsername/Heart-disease-predictor.git)

# Install dependencies
pip install -r requirements.txt

# Launch the application
streamlit run app/app.py











![alt text]({37630E2B-5050-48C1-8287-A30238E8BB4F}.png)
![alt text](image.png)
![alt text](image-2.png)