A web-based machine learning application that allows users to train and test models on both tabular (CSV) data and images without writing any code.

• Classification: Train models to categorize data into different classes
• Regression: Train models to predict continuous numerical values
• Feature Selection: Choose which columns to include in training
• Model Metrics: View accuracy (classification) or R² score/MSE (regression)
• Feature Importance: See which features contribute most to predictions

• Transfer Learning: Uses MobileNetV2 (pretrained on ImageNet) for feature extraction
• Multi-class Support: Upload multiple ZIP files, one per class (e.g., cats.zip, dogs.zip)
• Easy Testing: Upload new images to classify after training

• Frontend: React + TypeScript, Vite, Tailwind CSS
• Backend: Flask (Python)
• ML Models: scikit-learn (RandomForest)
• Deep Learning: TensorFlow/Keras (MobileNetV2 for image features)

1. Navigate to the backend directory:

   cd backend

2. Create a virtual environment (recommended):

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

3. Install dependencies:

   pip install -r requirements.txt

1. Navigate to the frontend directory:

   cd frontend

2. Install dependencies:

   npm install

cd backend
python app.py

The backend API runs on http://localhost:5000

cd frontend
npm run dev

The frontend runs on http://localhost:5173

python app.py

3. Open your browser and go to http://127.0.0.1:5000

1. Select "Tabular (CSV)" from the dropdown
2. Choose Classification or Regression mode
3. Upload your CSV file
4. Uncheck any columns that might cause data leakage
5. Enter the target column name
6. Click "Train CSV Model"

1. Select "Visual (Image ZIPs)" from the dropdown
2. Prepare ZIP files for each class (e.g., cats.zip containing cat images)
3. Upload all ZIP files at once
4. Click "Train Image Model"

After training, the prediction panel will update:

• CSV: Enter values for each feature and click "Predict"
• Images: Upload an image and click "Classify Image"