🧠 ML Playground

ML Playground is an intuitive, full-stack web application that empowers users to build, train, evaluate, visualize, and download machine learning models — all through an interactive UI and without writing a single line of code.

🔗 Live Demo

---

🚀 Tech Stack

• Backend: Django, Django REST Framework, PostgreSQL
• Frontend: React (Vite)
• Storage: Backblaze B2
• ML Libraries: scikit-learn, pandas, numpy, joblib

🎯 Purpose

The goal of ML Playground is to simplify the learning and experimentation process in machine learning. It serves as an educational tool where users can explore how different preprocessing techniques, algorithms, and parameter combinations influence model accuracy and performance — all visually and hands-on.

---

🧠 Key Features

• No Code ML: Upload CSV datasets and configure training pipelines with scalers, encoders, features, targets, and models via forms
• Data Preprocessing: Auto-handles missing values, outliers, and encoding internally
• Model Training: Select algorithm, adjust parameters, train the model, and view results instantly
• Model Management: Save, download (as .zip with all components), or delete trained models
• Live Predictions: Make real-time predictions with custom input on saved models
• Dashboard: Track models created, accuracy trends, and frequently used algorithms
• Learn Page: Embedded documentation to help beginners understand core ML concepts

🧪 Planned Enhancements

• AutoML recommendations based on dataset heuristics
• Team collaboration support
• Model version tracking and rollback
• Export to Jupyter notebooks
• Pipeline comparison charts

---

🗂️ Project Structure

Backend (backend/)

backend/
├── backend/                   # Django project configuration
│   ├── settings.py
│   └── urls.py
├── backend_app/
│   ├── users/
│   │   ├── views.py
│   │   ├── urls.py
│   │   └── serializers.py
│   ├── files/
│   │   ├── views.py
│   │   ├── urls.py
│   │   └── serializers.py
│   ├── models.py
│   └── admin.py
└── media/
    ├── uploads/
    ├── saved_models/
    ├── saved_encoders/
    ├── saved_scalers/
    └── saved_target_encoders/

Frontend (frontend/)

frontend/
└── src/
    ├── assets/
    ├── modals/
    ├── components/
    ├── pages/
    ├── styles/
    ├── App.jsx
    └── ProtectedRoute.jsx

---

📦 Backend Setup

1. Clone the repository:

   git clone <your-repo-url>
   cd backend

2. Create a virtual environment:

   python -m venv venv
   source venv/bin/activate  # Windows: venv\Scripts\activate

3. Install dependencies:

   pip install -r requirements.txt

4. Run migrations and create a superuser:

   python manage.py makemigrations
   python manage.py migrate
   python manage.py createsuperuser

5. Start the server:

   python manage.py runserver

💻 Frontend Setup

1. Navigate to frontend:

   cd frontend

2. Install dependencies:

   npm install

3. Run the app locally:

   npm run dev

---

🔐 Authentication Endpoints

Method	Endpoint	Description
POST	/api/token/	Get access & refresh token
POST	/api/token/refresh/	Refresh access token
POST	/login/	Custom login view
POST	/logout/	Logout
POST	/register/	User registration
GET	/profile/	Get user profile
GET	/secret-questions/	List of secret questions
POST	/verify-secret-answer/<username>/	Verify answer
POST	/password-reset-request/	Request password reset
POST	/reset-password/<username>/	Set new password
POST	/change-password/	Change password
GET	/user-secret-question/<username>/	Get user's secret question

📊 ML & Dataset Endpoints

Method	Endpoint	Description
POST	/upload/	Upload dataset
GET	/dataset-preview/	Preview dataset structure
POST	/train/	Train a model
POST	/save/	Save trained model
GET	/saved-model/<pk>/	Get model details
GET	/download-model/<pk>/	Download model ZIP
POST	/predict/<pk>/	Make prediction
GET	/dashboard-stats/	User dashboard data

---

👤 Author

• Name: Md Javed
• Email: [email protected]
• LinkedIn: www.linkedin.com/in/mdjaved077/

📄 License

This project is currently not licensed. You may use and modify it for educational or personal projects, but commercial use is not permitted unless granted explicit permission by the author.

---

✨ Built with passion to make Machine Learning accessible and intuitive ✨