Heart Disease Predictor API

A machine learning project to predict the presence of heart disease using clinical and demographic data, deployed using a FastAPI backend and containerized with Docker.

---

Overview

This project develops and deploys a machine learning model to predict the likelihood of heart disease based on patient clinical features such as age, cholesterol level, chest pain type, and others.

Multiple classification algorithms were trained and evaluated, including Logistic Regression, Support Vector Classifier (SVC), K-Nearest Neighbors (KNN), Decision Tree, Random Forest, AdaBoost, and XGBoost. After comparative model selection, SVC was chosen for its superior performance on the validation set.

The selected model is served via a RESTful API built with FastAPI. The entire application is containerized with Docker, enabling consistent and portable deployment across environments that support containers.

---

Project Structure

HeartPredictionAPI/
├── main.py               # FastAPI app
├── svm_pipeline.pkl      # Trained ML pipeline
├── model_training.ipynb  # Model development notebook (optional)
├── requirements.txt      # Python dependencies
├── Dockerfile            # Docker container instructions
├── README.md             # Project documentation
├── .gitignore
└── images/               # Screenshots for API demo

---

Dataset

The dataset is taken from Kaggle, which includes clinical features such as:

• Age, sex, chest pain type (cp), resting blood pressure (trestbps), cholesterol (chol)
• Fasting blood sugar (fbs), ECG (restecg), max heart rate (thalach)
• Exercise-induced angina (exang), ST depression (oldpeak), number of vessels (ca), and thalassemia type (thal)

---

Final Model

• Algorithm: Support Vector Machine (RBF Kernel)
• Pipeline: StandardScaler + SVM using Pipeline from scikit-learn
• Exported As: svm_pipeline.pkl using pickle
• Tuned With: GridSearchCV
• Evaluated Using: Accuracy, Precision, Recall, F1-Score, and SHAP for model explainability

---

Local Execution (via FastAPI)

1. Environment Setup

conda create -n heartapi python=3.12 -y
conda activate heartapi

2. Clone the Repository

git clone https://github.com/Amaan-developpeur/HeartPredictionAPI.git
cd HeartPredictionAPI

3. Install Dependencies

pip install -r requirements.txt

4. Run the API Server

uvicorn main:app --reload

Visit the interactive Swagger UI at:
📎 http://127.0.0.1:8000/docs

---

Sample API Request

{
  "age": 55,
  "sex": 1,
  "cp": 2,
  "trestbps": 130,
  "chol": 250,
  "fbs": 0,
  "restecg": 1,
  "thalach": 160,
  "exang": 0,
  "oldpeak": 1.5,
  "slope": 2,
  "ca": 0,
  "thal": 3
}

---

Docker Setup

You can run the entire app in a containerized environment using Docker.

1. Build the Docker Image

docker build -t heart-api .

2. Run the Container

docker run -p 8000:8000 heart-api

Then open your browser to:
📎 http://localhost:8000/docs

---

API Demo

🔹 Swagger UI Input Form

Users can submit patient data here:

---

🔹 API Response with Prediction

The model returns a clear message about heart disease risk:

---

🔹 Terminal Running FastAPI

Server running successfully with Uvicorn:

---

Requirements

See requirements.txt for exact versions. Core dependencies include:

• fastapi
• uvicorn
• pandas
• scikit-learn
• numpy
• python-multipart

---

---

Author

Mohammed Amaan
📎 GitHub: Amaan-developpeur