machine-learning-visualized

A Jupyter Book that implements machine learning algorithms in NumPy and derives the math from first principles. Meant for learners who want to understand how the algorithms work.

Browse open issues

Pick a small issue you can reproduce, then fix it with your AI coding agent.

Share on XLicense: MIT

Overview

Machine Learning Visualized is a Jupyter Book that teaches machine learning algorithms by deriving them from first principles in Jupyter Notebooks with NumPy. The output of each notebook is a visualization of the algorithm during training, ending at its optimal weights. There are also interactive notebooks built with Marimo that show how weights affect loss functions. This repository only holds the code that configures and builds the book, while each algorithm has its own separate repository.

Key features

  • Notebooks that derive algorithms mathematically with NumPy
  • Visualizes training until the weights converge
  • Interactive Marimo notebooks showing weights against loss
  • Builds a website from Markdown and notebooks with Jupyter Book

Best for

Learners who want to see how ML algorithms work step by step rather than call a library. The algorithm code itself lives in separate repositories.

Upstream
gavinkhung/machine-learning-visualized
Fork on GitHub
Guo-astro/machine-learning-visualized
Upstream stars
2.1k
Category
Learning, interviews and curated lists
Language
TeX
License
MIT
Forked
2026-08-11
Sync status
In syncLast synced 2026-10-20

More in Learning, interviews and curated lists