VisiML - Interactive Machine Learning Visualization Tool 🧠

VisiML is an interactive educational tool designed to help students and practitioners understand machine learning algorithms through real-time visualization. Watch how algorithms learn, adjust hyperparameters on the fly, and see the immediate impact on model performance.

🌟 Features

Supported Algorithms

Regression

• Linear Regression - Multiple optimizers (SGD, Adam, RMSprop), regularization options
• Polynomial Regression - Automatic feature engineering with degree control

Classification

• Logistic Regression - Binary and multi-class support
• Naive Bayes - Gaussian, Multinomial, and Bernoulli variants
• Decision Tree - Visual tree structure, feature importance
• Random Forest - Ensemble visualization, OOB score tracking
• Support Vector Machine - Kernel methods, margin visualization
• K-Nearest Neighbors - Distance metrics, Voronoi regions

Key Features

• 🎯 Real-time Training Visualization - Watch models learn iteration by iteration
• 🎛️ Interactive Hyperparameter Tuning - Adjust parameters and see immediate effects
• 📊 Comprehensive Metrics - Loss curves, accuracy, confusion matrices, ROC curves
• 🎨 Decision Boundary Visualization - See how models separate classes
• 📈 Performance Analysis - Learning curves, residual plots, feature importance
• 💾 Export Capabilities - Save models, plots, and generate code
• 🎲 Built-in Data Generation - Various patterns for testing (linear, spiral, moons, etc.)

🚀 Quick Start

Installation

1. Clone the repository:

git clone https://github.com/your-username/VisiML.git
cd VisiML

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

Running VisiML

Streamlit Web App (Recommended)

streamlit run app.py

Then open your browser to http://localhost:8501

Standalone Matplotlib Version

python visiml_standalone.py

📖 Usage Guide

Basic Workflow

1. Select Task Type: Choose between Regression or Classification
2. Choose Algorithm: Pick from available algorithms for your task
3. Configure Data:
   • Use built-in data generators with various patterns
   • Or upload your own CSV file
4. Adjust Hyperparameters: Use intuitive sliders and controls
5. Train & Visualize: Watch the model learn in real-time
6. Analyze Results: Explore metrics, visualizations, and insights

Example: Linear Regression

from visiml.models import LinearRegression
from visiml.data_generator import DataGenerator
from visiml.visualization import plot_regression_predictions

# Generate sample data
X, y = DataGenerator.generate_regression_data(
    n_samples=100, 
    function_type='polynomial',
    noise=0.1
)

# Create and train model
model = LinearRegression(learning_rate=0.01, n_iterations=1000)
model.fit(X, y)

# Visualize results
plot_regression_predictions(X, y, model)

📁 Project Structure

VisiML/
├── app.py                    # Streamlit web application
├── visiml_standalone.py      # Matplotlib standalone version
├── visiml/
│   ├── __init__.py
│   ├── models.py            # ML model implementations
│   ├── data_generator.py    # Data generation utilities
│   ├── visualization.py     # Visualization functions
│   └── utils.py            # Helper functions
├── examples/
│   ├── linear_regression.ipynb
│   ├── classification_demo.ipynb
│   └── custom_data_example.ipynb
├── docs/
│   ├── user_guide.md
│   ├── api_reference.md
│   └── images/
├── tests/
│   ├── test_models.py
│   ├── test_visualization.py
│   └── test_data_generator.py
├── requirements.txt
├── setup.py
├── LICENSE
└── README.md

🎓 Educational Use Cases

VisiML is perfect for:

• Machine Learning Courses - Interactive demonstrations in lectures
• Self-Study - Hands-on learning of ML concepts
• Algorithm Comparison - Side-by-side algorithm performance
• Hyperparameter Understanding - See effects of different parameters
• Debugging Models - Visualize what's happening inside algorithms

🛠️ Advanced Features

Custom Data Upload

Support for CSV files with automatic feature and target detection.

Model Export

• Save trained models as pickle files
• Export training history and metrics
• Generate Python code for reproduction

Visualization Options

• 2D and 3D plotting capabilities
• Animation of training progress
• Interactive plots with hover information

🤝 Contributing

We welcome contributions! Please see our Contributing Guidelines for details.

Development Setup

1. Fork the repository
2. Create a feature branch (git checkout -b feature/AmazingFeature)
3. Commit changes (git commit -m 'Add AmazingFeature')
4. Push to branch (git push origin feature/AmazingFeature)
5. Open a Pull Request

📄 License

This project is licensed under the MIT License - see the LICENSE file for details.

🙏 Acknowledgments

• Inspired by the need for better ML education tools
• Built with Streamlit, Matplotlib, and scikit-learn
• Thanks to all contributors and users

🚧 Roadmap

• Add deep learning models (Neural Networks)
• Support for clustering algorithms
• Time series visualization
• Model comparison dashboard
• Export to TensorBoard
• Mobile-responsive design

---

Made with ❤️ for ML Education