Stroke Prediction Model

This project aims to predict the likelihood of a stroke using various machine learning models based on the healthcare-dataset-stroke-data.csv dataset. The dataset contains information on various health attributes that may contribute to stroke occurrence.

Table of Contents

• Installation
• Usage
• Dataset
• Data Preprocessing
• Model Building
• Results
• Contributing
• License

Installation

1. Clone the repository:

   git clone https://github.com/Elilora/stroke-prediction.git
   cd stroke-prediction

2. Install the required libraries:

   pip install numpy pandas seaborn matplotlib scikit-learn xgboost shap

Usage

1. Ensure the dataset file healthcare-dataset-stroke-data.csv is in the /kaggle/input/stroke-prediction-dataset/ directory.

2. Run the script to train and evaluate the models:

   python stroke_prediction.py

Dataset

The dataset used for this project is healthcare-dataset-stroke-data.csv, which contains the following columns:

• id
• gender
• age
• hypertension
• heart_disease
• ever_married
• work_type
• Residence_type
• avg_glucose_level
• bmi
• smoking_status
• stroke

Data Preprocessing

1. Handle missing values:
   • The bmi column had missing values, which were imputed using the mean strategy.
2. Encode categorical variables:
   • Categorical columns were encoded using LabelEncoder.

Model Building

The following models were used to predict stroke occurrence:

• Decision Tree Classifier
• Random Forest Classifier

Steps

1. Split the data into training and testing sets (70% train, 30% test).
2. Train the models on the training set.
3. Evaluate the models on the testing set using accuracy and F1-score metrics.
4. Generate classification reports and confusion matrices.

Results

Decision Tree Classifier

• Accuracy: 91.06%
• F1-score: 10.46%
• Detailed classification report and confusion matrix are generated in the script output.

Random Forest Classifier

• Accuracy: 95.50%
• F1-score: 0.00%
• Detailed classification report and confusion matrix are generated in the script output.

Contributing

Contributions are welcome! Please fork the repository and submit a pull request for any enhancements or bug fixes.

License

This project is licensed under the MIT License - see the LICENSE file for details.