🎓 Supervised Machine Learning

This repository teaches machine learning from first principles using Python.

Starting with foundational mathematics (derivatives, exp/log, probability), each notebook builds complete understanding before exploring modern tools like Scikit-learn and PyTorch. All mathematical concepts are derived step-by-step, making the content accessible to anyone with basic high school calculus knowledge.

This aims to be part of a broader machine learning series, with companion repositories for unsupervised, reinforcement and other state-of-the-art machine learning domains in development.

📚 Notebooks

🚀 Quick Start: Run notebooks directly in your browser - no setup required!

1. Click any "Open in Colab" button below
2. In Colab: Click "Connect" (top-right)
3. Click "Runtime" > "Run all" (top menu)

1a_logistic_regression_theory.ipynb

Theory & from-scratch implementation of logistic regression using the Wisconsin Breast Cancer dataset.

• 
• View Source

1b_logistic_regression_practical.ipynb

Production-grade PyTorch implementation with modern ML engineering practices.

• 
• View Source

2a_decision_trees_theory.ipynb

Deep dive into decision tree theory with a complete from-scratch implementation.

• 
• View Source

2b_decision_trees_practical.ipynb

Real-world prediction of London housing market prices with Scikit-learn, random forests and XGBoost.

• 
• View Source

2c_decision_trees_ATLAS_model_comparison.ipynb

Automated Tree Learning Analysis System (ATLAS) for feature engineering and model comparison.

• 
• View Source

🧠 Neural Networks (Coming Soon)

Implementation of neural networks and deep learning fundamentals.

📊 Datasets

Wisconsin Breast Cancer (1995)

• Binary classification task
• 569 samples × 30 features
• Medical diagnosis application

London Housing Prices (2020)

• Regression problem
• 3,479 samples × 9 features
• Geographic feature encoding

💻 Local Setup

For those who prefer to run notebooks locally:

git clone https://github.com/powell-clark/supervised-machine-learning.git
cd supervised-machine-learning
python3 -m venv venv
source venv/bin/activate
pip install -r requirements.txt

Contributing & License

Copyright © 2025 Powell-Clark Limited
Licensed under the Apache License, Version 2.0

Contributing

We welcome community contributions to improve these educational materials! Here's how you can help:

• Spot a typo or unclear explanation? Open a pull request
• Have an idea for improvement? Create an issue
• Want to add new examples or exercises? We'd love to see them
• Found a bug in the code? Let us know

Contributors will be acknowledged in our CONTRIBUTORS.md file which will be created with our first contribution.

Citation

If you use these materials in your work, please cite as:

Powell-Clark (2025). Supervised Machine Learning from First Principles.
GitHub: https://github.com/powell-clark/supervised-machine-learning

Contact

Questions or feedback? Contact [email protected]