Heart Disease Prediction Model using Logistic Regression

A machine learning project that predicts the presence or absence of heart disease in patients using medical diagnostic features and logistic regression classification.

📋 Table of Contents

• Overview
• Dataset
• Features
• Model
• Results
• Installation
• Usage
• Project Structure
• Technologies Used
• Contributing
• License

🎯 Overview

This project develops a machine learning model to predict heart disease using logistic regression. The model is designed to serve as a clinical decision support tool that helps identify patients at risk of heart disease based on various medical diagnostic measurements. The project emphasizes achieving a balance between accuracy, precision, and recall for reliable healthcare applications.

Key Highlights:

• ✅ Achieves 86% accuracy on test data
• ✅ Excellent AUC of 0.92 demonstrating strong discriminative ability
• ✅ Uses hyperparameter tuning for optimal performance
• ✅ Implements feature standardization for better model convergence
• ✅ Provides comprehensive model evaluation metrics

📊 Dataset

The project uses a preprocessed heart disease dataset containing 303 patient records. The data is split using:

• Training set: 70% of the data (stratified sampling)
• Testing set: 30% of the data (stratified sampling)

Stratified sampling ensures that the class distribution (disease vs. no disease) is maintained in both training and testing sets.

🔍 Features

The model uses 13 predictor variables to make predictions:

Feature	Description	Type
Age	Patient's age in years	Continuous
Sex	Gender (1 = male, 0 = female)	Binary
Chest Pain Type	Type of chest pain (1-4)	Categorical
BP	Resting blood pressure (mm Hg)	Continuous
Cholesterol	Serum cholesterol level (mg/dl)	Continuous
FBS over 120	Fasting blood sugar > 120 mg/dl	Binary
EKG Results	Resting electrocardiographic results	Categorical
Max HR	Maximum heart rate achieved	Continuous
Exercise Angina	Exercise-induced angina	Binary
ST Depression	ST depression induced by exercise	Continuous
Slope of ST	Slope of the peak exercise ST segment	Categorical
Number of Vessels Fluro	Number of major vessels colored by fluoroscopy (0-3)	Categorical
Thallium	Thallium stress test results	Categorical

Target Variable: Presence (1) or Absence (0) of heart disease

🤖 Model

Algorithm: Logistic Regression

The project implements logistic regression with several optimization techniques:

1. Feature Standardization: StandardScaler is applied to normalize all features, ensuring they have zero mean and unit variance
2. Hyperparameter Tuning: GridSearchCV with 5-fold cross-validation to find optimal parameters
3. Optimization Metric: F1-score (balances precision and recall)
4. Model Evaluation: Comprehensive metrics including confusion matrix, classification report, ROC curve, and AUC

Model Pipeline

Raw Data → Feature Scaling → Train-Test Split → Logistic Regression → Hyperparameter Tuning → Final Model

📈 Results

Best Model Performance (Hyperparameter-Tuned)

Metric	Score
Accuracy	86%
Precision (Disease)	87%
Recall (Disease)	79%
F1-Score (Disease)	83%
AUC-ROC	0.92

Model Comparison

Model	Accuracy	Recall	Precision	F1-Score
Baseline	84%	76%	87%	81%
Tuned (Final)	86%	79%	87%	83%

The tuned model shows improvement across all key metrics while maintaining excellent discriminative ability (AUC = 0.92), making it suitable for clinical decision support.

🚀 Installation

Prerequisites

• Python 3.8+ (Python 3.9 or higher recommended)
• Google Colab (optional, for running the notebook)
• Jupyter Notebook (for local execution)

Required Libraries

pip install pandas numpy scikit-learn matplotlib seaborn jupyter

Or install with specific tested versions from requirements.txt:

pip install -r requirements.txt

💻 Usage

Running in Google Colab

1. Open the notebook: Heart Disease Prediction Model using Logistic Regression.ipynb
2. Mount Google Drive (if your dataset is stored there)
3. Update the file_path variable to point to your preprocessed dataset
4. Run all cells sequentially

Running Locally

1. Clone this repository:

git clone https://github.com/DataDarling/Heart-Disease-Prediction-Model.git
cd Heart-Disease-Prediction-Model

2. Install dependencies:

pip install -r requirements.txt

3. Launch Jupyter Notebook:

jupyter notebook "Heart Disease Prediction Model using Logistic Regression.ipynb"

4. Update the dataset path in the notebook and run all cells

Using the Model for Predictions

import pandas as pd
from sklearn.preprocessing import StandardScaler
from sklearn.linear_model import LogisticRegression

# Load your trained model
# ... (train the model as shown in the notebook)

# Prepare new patient data
new_patient = [[63, 1, 3, 145, 233, 1, 0, 150, 0, 2.3, 0, 0, 1]]

# Scale the features
new_patient_scaled = scaler.transform(new_patient)

# Make prediction
prediction = model.predict(new_patient_scaled)
probability = model.predict_proba(new_patient_scaled)

print(f"Prediction: {'Heart Disease' if prediction[0] == 1 else 'No Heart Disease'}")
print(f"Probability: {probability[0][1]:.2%}")

📁 Project Structure

Heart-Disease-Prediction-Model/
│
├── Heart Disease Prediction Model using Logistic Regression.ipynb
│   └── Main notebook with complete analysis and model development
│
└── README.md
    └── Project documentation (this file)

🛠️ Technologies Used

• Python 3.8+ - Primary programming language
• Pandas - Data manipulation and analysis
• NumPy - Numerical computing
• Scikit-learn - Machine learning library
  • LogisticRegression
  • StandardScaler
  • GridSearchCV
  • train_test_split
  • Classification metrics
• Matplotlib - Data visualization
• Seaborn - Statistical data visualization
• Google Colab - Development environment

🤝 Contributing

Contributions are welcome! Here's how you can help:

1. Fork the repository
2. Create a feature branch (git checkout -b feature/improvement)
3. Commit your changes (git commit -m 'Add some improvement')
4. Push to the branch (git push origin feature/improvement)
5. Open a Pull Request

Ideas for Contribution

• Try different algorithms (Random Forest, SVM, Neural Networks)
• Add feature engineering techniques
• Implement cross-validation strategies
• Create a web interface for predictions
• Add more visualizations
• Improve documentation

📄 License

This project is open source and available under the MIT License.

📞 Contact

For questions or feedback, please open an issue in the GitHub repository.

---

Note: This model is for educational and research purposes only. It should not be used as a substitute for professional medical advice, diagnosis, or treatment. Always consult with qualified healthcare providers for medical decisions.