About ML Visualization
Mission
ML Visualization is a free, interactive encyclopedia of machine learning. Every concept here is something you can touch: drag the data, move the parameters, break the model, and watch the math respond. We believe you understand an algorithm when you can play with it — not when you’ve watched an animation play at you.
Who it’s for
Curious beginners, students, teachers, and working engineers who want intuition fast. No sign-up, no cost, no prerequisites beyond curiosity.
How to use it
Head to the Visualizations gallery and pick anything that looks interesting. Each page pairs a live visualization with the relevant formulas — every symbol explained — and a “now break it” section. If you’d like a guided order, follow the Learning Path.
Scope
This is a classical machine learning encyclopedia — foundations, regression, classification, ensembles, unsupervised learning, evaluation, and neural networks through backpropagation. That range is a deliberate choice, and it’s covered end to end: every concept in it is built, not stubbed. Deep learning and probability & statistics are their own subjects with their own intuitions, so they sit outside this scope rather than getting a thin corner here — a complete map of one territory beats a sketch of three.
How it’s built
Every visualization computes in your browser, not on a server — the math runs on your device. We don’t ask you to sign up or fill in forms. The site is supported by ads; for how advertising and cookies work, see our Privacy Policy. The math is typeset for clarity and content is reviewed for accuracy — found a mistake? We’d genuinely like to know.
Contact
Found a bug, an error in the math, or have a concept you’d love to see? Email us at mlvisualizationorg@gmail.com — we read everything.