Heart Disease Prediction Model

A machine learning project that predicts heart disease using clinical features from the UCI Cleveland Heart Disease Dataset. The model is built with a Random Forest Classifier and optimized using Scikit-learn + GridSearchCV for hyperparameter tuning.

Features

This model uses 13 clinical features to predict heart disease:

Feature	Description	Range
age	Age of the patient in years	20-100
sex	Sex of the patient (0 = female, 1 = male)	0-1
cp	Chest pain type (1-4)	1-4
trestbps	Resting blood pressure (in mm Hg)	80-200
chol	Serum cholesterol (in mg/dl)	100-600
fbs	Fasting blood sugar > 120 mg/dl (0 = false, 1 = true)	0-1
restecg	Resting electrocardiographic results	0-2
thalach	Maximum heart rate achieved	60-220
exang	Exercise-induced angina (0 = no, 1 = yes)	0-1
oldpeak	ST depression induced by exercise relative to rest	0-6.2
slope	Slope of the peak exercise ST segment	1-3
ca	Number of major vessels colored by fluoroscopy	0-3
thal	Thalassemia (3 = normal, 6 = fixed defect, 7 = reversible defect)	3-7

Dataset

The model uses the UCI Cleveland Heart Disease Dataset from the UCI Machine Learning Repository.

• Source: UCI Machine Learning Repository
• Records: 303 patient records
• Target Variable: Binary classification (0 = no heart disease, 1 = heart disease)

Class Distribution

• No heart disease (0): ~54%
• Heart disease (1): ~46%

Model

Algorithm

• Random Forest Classifier: An ensemble learning method that operates by constructing a multitude of decision trees at training time.

Hyperparameter Optimization

• Method: GridSearchCV with 5-fold cross-validation
• Parameters Optimized:
  • n_estimators: [100, 200, 300, 400, 500]
  • max_depth: [None, 10, 20, 30]
  • min_samples_split: [2, 5, 10]
  • min_samples_leaf: [1, 2, 4]
  • bootstrap: [True, False]

Installation

Prerequisites

• Python 3.8 or higher
• pip (Python package installer)

Dependencies

Install the required packages:

pip install pandas numpy matplotlib scikit-learn joblib

Or install all dependencies from requirements.txt:

pip install -r requirements.txt

Usage

Training the Model

Run the main training script to:

1. Download and preprocess the dataset
2. Train and optimize the Random Forest model
3. Generate feature importance visualization
4. Save the model, scaler, and feature names

python main.py

Making Predictions

Use the prediction CLI tool to input patient data and get heart disease predictions:

python heart_disease_predictor.py

The tool will prompt for each clinical feature with validation against expected ranges. Enter values or type 'exit' to quit.

Example session:

Heart Disease Prediction Tool
----------------------------
Loaded feature names: ['age', 'sex', 'cp', 'trestbps', 'chol', 'fbs', 'restecg', 'thalach', 'exang', 'oldpeak', 'slope', 'ca', 'thal']
Number of features: 13

Enter the patient's data (or type 'exit' to quit):
Age of the patient in years (expected range: (20, 100)): 55
Sex of the patient (0 = female, 1 = male) (expected range: (0, 1)): 1
...

Project Structure

Heart disease prediction model/
├── main.py                          # Main training script
├── heart_disease_predictor.py       # Prediction CLI tool
├── features.py                      # Feature definitions and ranges
├── readme.md                        # This file
├── .gitignore                       # Git ignore rules
├── .vscode/                         # VSCode settings
├── 9000/                            # Additional resources
├── data.json                        # Project data
├── new_data.csv                     # Processed dataset
├── prediction_log.csv               # Prediction history log
├── feature_importance.png           # Feature importance visualization
├── feature_names.pkl                # Feature names pickle
├── scaler.pkl                       # StandardScaler pickle
├── optimized_heart_disease_model.pkl # Trained model pickle
└── heart_disease_model.pkl          # Original model pickle

Output Files

File	Description
optimized_heart_disease_model.pkl	Trained Random Forest model with optimized hyperparameters
scaler.pkl	StandardScaler fitted on training data for feature normalization
feature_names.pkl	List of feature names used by the model
feature_importance.png	Bar chart showing feature importance rankings
prediction_log.csv	CSV file logging all predictions with timestamps, patient IDs, input values, predictions, and probabilities

Data Pipeline

Preprocessing Steps

1. Data Loading: Download dataset from UCI repository
2. Missing Value Handling: Replace '?' with NaN, then fill with mode
3. Type Conversion: Convert 'ca' and 'thal' columns to numeric types
4. Target Binarization: Convert target values to binary (0 or 1)
   • Original values > 0 become 1 (heart disease present)
   • Original values = 0 become 0 (no heart disease)
5. Feature Scaling: StandardScaler normalization (mean=0, std=1)
6. Train-Test Split: 80% training, 20% testing (random_state=42)

Pipeline Diagram

UCI Dataset → Missing Value Handling → Type Conversion → Target Binarization
                                                                    ↓
Feature Scaling → Train-Test Split → GridSearchCV → Optimized Model

Evaluation Metrics

The model is evaluated on the test set using the following metrics:

Metric	Description	Target
Accuracy	Proportion of correct predictions	> 80%
Precision	True positives / (True positives + False positives)	> 80%
Recall	True positives / (True positives + False negatives)	> 80%
ROC-AUC	Area under the ROC curve	> 85%

Sample Performance (Test Set)

Accuracy:  0.85
Precision: 0.84
Recall:    0.86
ROC-AUC:   0.92

Feature Importance

The model identifies the most important features for heart disease prediction. Common top features include:

1. ca - Number of major vessels
2. thal - Thalassemia test result
3. oldpeak - ST depression
4. thalach - Maximum heart rate
5. cp - Chest pain type

A feature importance visualization is saved as feature_importance.png after training.

Contributing

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

License

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

Acknowledgments

• UCI Machine Learning Repository for the Heart Disease Dataset
• scikit-learn for the machine learning tools