Course 0 — Warm-Up

If you've never touched ML, math, or code — start here. You don't need to be "good at math" or to have written a line of code before. At the start, machine learning is concrete: one model, one line through some points, one number going down as the model learns. This is the easier, more rewarding place to begin.

Everything here runs in your browser. No install, no terminal, no setup. If you can edit a number and read a few short lines of Python, you can do every lesson — no prior ML and no math background needed.

What you'll do

Seven short modules, about an hour end to end:

1. Your first model — run real Python, change one number, watch a prediction move.
2. Become the algorithm — fit a line to points by hand, then meet the algorithm that does it for you.
3. The only calculus you need — slope, as a picture you can drag. (Single variable — that's all you need here.)
4. NumPy without fear — the eight operations the real courses actually use.
5. Vectors, dot products & layers — the language ML is written in.
6. You've already started — train a model end to end and watch the loss fall.
7. How do we know it's any good? — does the model work on data it never saw, plus the one-liner that does it for you.

By the end you'll have trained a model yourself and read real Python — and you'll be ready for Foundations of Regression, the first full course.