Heart_Disease_Prediction

❤️ Heart Disease Prediction App

A machine learning web app that predicts heart disease risk using patient health data. Built with Streamlit for the interface and scikit-learn for the models.

Overview

This project trains four different machine learning models on patient data, compares their performance, and uses the most accurate one to make predictions through an interactive web interface. Users can either:

• Enter a single patient's details and get an instant prediction, or
• Upload a CSV file to get predictions for many patients at once.

Models Used

The app runs new input through all four models internally, and displays the final prediction from whichever model has the highest test accuracy.

Project Structure

.
├── app.py                    # Streamlit web app (3 tabs: Predict, Bulk Predict, Model Information)
├── decision.pkl               # Trains the Decision Tree model
├── logistic.pkl                # Trains the Logistic Regression model
├── support.pkl                # Trains the SVM model
├── random.pkl                # Trains the Random Forest model, scaler, and feature list
├── requirements.txt            # Python dependencies
├── sample_bulk_predict.csv     # Example CSV for testing the Bulk Predict tab
└── README.md

Note: The training scripts use a .pkl extension (despite being Python scripts) to match the original project naming. They are plain Python files and should be run with python <filename>.pkl.

Dataset

This project uses the Statlog Heart Disease dataset format, with the following columns:

Age, Sex, Chest pain type, BP, Cholesterol, FBS over 120, EKG results,
Max HR, Exercise angina, ST depression, Slope of ST,
Number of vessels fluro, Thallium, Heart Disease

The target column Heart Disease contains Presence / Absence labels, which are mapped to 1 / 0 during training.

The dataset file (cleaned_data.csv) is not included in this repository. Place your own copy in the project root before running the training scripts.

Setup

1. Clone the repository

   git clone https://github.com/<your-username>/<your-repo>.git
   cd <your-repo>
   

2. Install dependencies

   pip install -r requirements.txt
   

3. Add your dataset Place cleaned_data.csv in the project root.

4. Train the models Run each training script once. This generates the model files (.pkl) the app needs:

   python decision.pkl
   python logistic.pkl
   python support.pkl
   python random.pkl
   

5. Run the app

   streamlit run app.py
   

   This opens the app in your browser at http://localhost:8501.

Feature Encoding Reference

App Features

• Predict tab — fill in a form with patient details and get an instant High Risk / Low Risk prediction with a confidence percentage.
• Bulk Predict tab — upload a CSV of multiple patients and download predictions for all of them at once.
• Model Information tab — view which model is currently used, its hyperparameters, and accuracy comparison across all four models.

Disclaimer

This app is for educational purposes only and is not a substitute for professional medical advice, diagnosis, or treatment. Always consult a qualified healthcare provider regarding any medical condition.