Gender Classifier Web Application

A full-stack web application that predicts a user’s gender based on biometric inputs using an ensemble of machine learning models. The application features a React frontend, a Flask REST API backend, and a PostgreSQL database for persistent training data.

---

🚀 Features

• Full-stack architecture using React, Flask, and PostgreSQL
• Machine learning ensemble approach with multiple classifiers:
  • K-Nearest Neighbors (KNN)
  • Quadratic Discriminant Analysis (QDA)
  • Random Forest
• Confidence-based prediction selection across models
• RESTful API for predictions and data ingestion
• Dynamic model training from database-stored user data
• User feedback loop to improve future predictions
• Cloud-ready backend configuration
• Interactive UI with built-in unit conversion guides

---

🧠 How It Works

1. Users input:
   • Height (cm)
   • Weight (kg)
   • Shoe size (EU)
2. The React frontend sends the data to the Flask API.
3. The backend:
   • Retrieves labeled training data from PostgreSQL
   • Trains multiple machine learning models
   • Generates predictions and confidence scores
   • Selects the prediction with the highest confidence
4. Results are returned to the frontend and displayed to the user.
5. Users can optionally submit feedback to improve future predictions.

---

🏗️ Tech Stack

Frontend

• React
• JavaScript (ES6+)
• CSS
• Fetch API

Backend

• Python
• Flask
• Flask-CORS
• SQLAlchemy
• scikit-learn

Database

• PostgreSQL

Machine Learning

• K-Nearest Neighbors
• Quadratic Discriminant Analysis
• Random Forest Classifier

---

📂 Project Structure

GenderClassifierWebApp/
├── api/
│   └── index.py          # Flask backend and ML logic
├── frontend/
│   ├── src/
│   │   ├── App.js        # Main React application
│   │   ├── App.css
│   │   └── assets/       # Conversion images
│   └── package.json
├── README.md