🫀 Swasya

🩺 Smart Heart Disease Risk Assessment using Machine Learning

Predict heart disease risk using Machine Learning and interactive data-driven insights.

---

Live Demo

🔗 Try the application here:

https://swasya.vercel.app

---

📌 Project Overview

Swasya is a Machine Learning-powered healthcare application designed to estimate the probability of heart disease using clinical and lifestyle parameters.

The application utilizes a Support Vector Classifier (SVC) model with probability estimation enabled, allowing users to receive risk scores rather than simple binary predictions.

The project combines data preprocessing, feature engineering, machine learning, and an interactive Streamlit dashboard to create an intuitive cardiovascular risk assessment tool.

⚠️ Disclaimer: This application is intended for educational and research purposes only and should not be considered a substitute for professional medical advice.

---

Application Preview

Home Dashboard

Prediction Results

---

Features

Interactive Streamlit Dashboard

Heart Disease Risk Prediction

Probability-Based Risk Assessment

Personalized Health Recommendations

Modern Dark-Themed User Interface

Interactive Sliders and Dropdown Controls

Health Awareness Tips Sidebar

Real-Time Predictions

Responsive Design

Confidence Score Visualization

---

Dataset Information

The model was trained on a heart disease dataset containing:

Attribute	Value
Records	918
Features	11
Target Variable	HeartDisease

Features Used

• Age
• Sex
• ChestPainType
• RestingBP
• Cholesterol
• FastingBS
• RestingECG
• MaxHR
• ExerciseAngina
• Oldpeak
• ST_Slope

---

Data Preprocessing

The following preprocessing steps were performed:

• Data Cleaning
• Handling Missing Values
• Categorical Feature Encoding
• Feature Scaling using StandardScaler
• Train-Test Split (80:20)
• Random State = 42

---

Machine Learning Models Evaluated

Multiple machine learning algorithms were trained and compared.

Model	Accuracy
Logistic Regression	86.41%
Support Vector Classifier (SVC)	85.33%
K-Nearest Neighbors (KNN)	85.33%
Gaussian Naive Bayes	84.78%
Decision Tree Classifier	78.80%

---

Final Model

The deployed application uses:

from sklearn.svm import SVC

model = SVC(
    kernel='linear',
    random_state=42,
    probability=True
)

Why SVC?

Although Logistic Regression achieved the highest accuracy, SVC was selected because probability estimation was enabled (probability=True).

This allows the application to provide:

• Risk Percentage
• Healthy Probability
• Confidence-Based Predictions
• Better User Experience

rather than only returning a binary output.

---

Technology Stack

Programming Language

• Python

Libraries

• Pandas
• NumPy
• Scikit-Learn
• Plotly
• Joblib

Framework

• Streamlit

---

Machine Learning Pipeline

Dataset
   ↓
Data Cleaning
   ↓
Feature Encoding
   ↓
Train-Test Split
   ↓
Feature Scaling
   ↓
Model Training
   ↓
SVC Classification
   ↓
Probability Prediction
   ↓
Streamlit Deployment

---

Project Structure

heart-disease-prediction/
│
├── app.py
├── heart.csv
├── heart_attack.ipynb
├── SVC_model.joblib
├── scaler.joblib
├── columns.joblib
├── requirements.txt
├── README.md
│
└── assets/
    ├── home.png
    └── prediction.png

---

Installation

Clone Repository

git clone https://github.com/supremeinferno/heart-disease-prediction.git

Navigate to Project Directory

cd heart-disease-prediction

Install Dependencies

pip install -r requirements.txt

Run Streamlit Application

streamlit run app.py

---

Future Improvements

• Explainable AI using SHAP
• Feature Importance Visualization
• Deep Learning Models
• User Authentication
• Medical Report Export
• Cloud Database Integration
• Mobile-Friendly Dashboard

---

👨‍💻 Author

Pranav Garg

Passionate about:

• Machine Learning • Computer Vision • Deep Learning • Agentic AI

🔗 GitHub: https://github.com/supremeinferno