https://github.com/pjaiswalusf/stroke-prediction
This project leverages machine learning to predict stroke risk using XGBoost, Random Forest, and Logistic Regression. It incorporates advanced data preprocessing, class imbalance handling with SMOTE, and hyperparameter optimization using Optuna. Model interpretability is enhanced with SHAP to identify key risk factors.
https://github.com/pjaiswalusf/stroke-prediction
data-science datapreprocessing logistic-regression machine-learning optuna random-forest shap smote xgboost
Last synced: 2 months ago
JSON representation
This project leverages machine learning to predict stroke risk using XGBoost, Random Forest, and Logistic Regression. It incorporates advanced data preprocessing, class imbalance handling with SMOTE, and hyperparameter optimization using Optuna. Model interpretability is enhanced with SHAP to identify key risk factors.
- Host: GitHub
- URL: https://github.com/pjaiswalusf/stroke-prediction
- Owner: pjaiswalusf
- Created: 2025-02-17T17:47:58.000Z (2 months ago)
- Default Branch: main
- Last Pushed: 2025-02-20T17:12:15.000Z (2 months ago)
- Last Synced: 2025-02-20T17:24:54.896Z (2 months ago)
- Topics: data-science, datapreprocessing, logistic-regression, machine-learning, optuna, random-forest, shap, smote, xgboost
- Language: Jupyter Notebook
- Homepage:
- Size: 3.51 MB
- Stars: 0
- Watchers: 1
- Forks: 0
- Open Issues: 0
-
Metadata Files:
- Readme: README.md
Awesome Lists containing this project
README
# Optimized Stroke Prediction with Machine Learning 🚀
## 📌 Overview
This project aims to predict the likelihood of stroke using **machine learning models** such as **XGBoost, Random Forest, and Logistic Regression**. The dataset consists of patient health records with various risk factors, and the model is optimized using **Optuna** for hyperparameter tuning. Additionally, **SHAP (Shapley Additive Explanations)** is utilized to enhance model interpretability by identifying the top contributing features.
## 🛠️ Technologies Used
- **Python** (Core programming language)
- **Scikit-learn** (Machine Learning Library)
- **XGBoost** (Boosted Decision Trees)
- **Random Forest** (Ensemble Learning)
- **Optuna** (Hyperparameter Tuning)
- **SHAP** (Explainable AI)
- **SMOTE** (Handling Class Imbalance)
- **Pandas & NumPy** (Data Manipulation)
- **Matplotlib & Seaborn** (Visualization)
## 📊 Dataset
- The dataset used in this project contains **5,110 patient records** with features like **age, hypertension, heart disease, smoking status, glucose level, BMI, and stroke occurrence**.
- Data Source: [Healthcare Stroke Prediction Dataset](https://www.kaggle.com/fedesoriano/stroke-prediction-dataset)
## 🔥 Key Features
✅ **Data Preprocessing & Feature Engineering**: Missing value imputation, outlier detection, encoding, and scaling.
✅ **Class Imbalance Handling**: Augmented minority class representation using **SMOTE** to balance stroke cases.
✅ **Model Optimization & Evaluation**: Fine-tuned **Random Forest & XGBoost**, achieving **97.4% accuracy**.
✅ **Explainable AI**: Used **SHAP** to highlight the **top 5 stroke risk factors**.
✅ **Hyperparameter Tuning**: Optimized model performance using **Optuna**, improving prediction precision and recall.
## ⚙️ Installation
To run this project locally, follow these steps:
1. **Clone the repository**:
```bash
git clone https://github.com/pjaiswalusf/Stroke-Prediction
cd stroke-prediction-ml
```
2. **Create and activate a virtual environment (optional but recommended):**
```bash
python -m venv venv # Create virtual environment
source venv/bin/activate # For Mac/Linux
venv\Scripts\activate # For Windows
```
3. **Install dependencies:**
```bash
pip install -r requirements.txt
```
4. **Run the Jupyter Notebook or Python script:**
```bash
jupyter notebook
```
Open `stroke_prediction.ipynb` and execute the cells.
## 📌 Usage
- **Exploratory Data Analysis (EDA):** Run the provided notebook to generate insights and visualize key features.
- **Train the Model:** Train different models using the dataset and compare performance metrics.
- **Hyperparameter Tuning:** Fine-tune the models using Optuna for the best results.
- **Explainability:** Use SHAP to analyze feature importance and understand model predictions.
## 📈 Model Performance
| Model | Accuracy | Precision | Recall | F1-score |
|--------|------------|------------|---------|----------|
| **XGBoost** | **97.4%** | 95.8% | 96.1% | 96.0% |
| **Random Forest** | 96.7% | 94.2% | 95.0% | 94.6% |
| **Logistic Regression** | 85.6% | 81.0% | 83.2% | 82.1% |
## 📊 Feature Importance (Top 5 Features by SHAP Analysis)
1️⃣ **Age** 🡆 Higher age increases stroke risk.
2️⃣ **Hypertension** 🡆 People with hypertension have a significantly higher probability of stroke.
3️⃣ **Heart Disease** 🡆 Patients with pre-existing heart disease are more likely to have a stroke.
4️⃣ **Glucose Level** 🡆 Elevated blood sugar levels strongly correlate with stroke occurrences.
5️⃣ **BMI** 🡆 Obesity is a contributing factor to stroke risk.
## 📌 Future Improvements
🚀 **Deploy the model as an API** using **Flask/FastAPI** for real-time stroke prediction.
🚀 **Improve class balancing techniques** to handle data skewness more effectively.
🚀 **Try deep learning models** (e.g., LSTMs or Neural Networks) for better accuracy.
## 🤝 Contributing
Feel free to **fork** this repository, **open issues**, or submit **pull requests**. Contributions are always welcome!
## 📜 License
This project is **open-source** and available under the **MIT License**.
## 📬 Contact
For any questions or collaborations, reach out to me at **[email protected]** or connect via [LinkedIn](https://linkedin.com/in/yourprofile](https://www.linkedin.com/in/pratik-jaiswal-468315197/)). 🚀