ML for Beginners exists to explain the mathematics behind machine learning to people who are meeting it for the first time.
Who writes this
My name is Jamal Hussein. I hold a BSc in Computer Technology and I am currently studying for an MSc in Artificial Intelligence.
I started this site out of frustration. When I began learning machine learning properly, almost every explanation of the underlying maths fell into one of two camps: a formal textbook treatment that assumed a mathematics degree, or a hand-waving summary that told me a gradient “points downhill” without ever showing me how to compute one. Neither is much use when you are staring at a loss function and cannot follow the derivation.
So I write the explanation I wanted: the actual working, step by step, in plain English, with the intermediate lines left in rather than skipped.
What is on the site
The material covers the three areas of mathematics that machine learning genuinely depends on, plus the ML concepts built on top of them:
- Linear algebra — vectors, matrices, eigenvalues; how data is represented
- Calculus — limits, derivatives, the chain rule; how models improve
- Probability and statistics — how you judge whether a result means anything
- Machine learning — loss functions, activation functions, evaluation metrics
- Calculators and tools — free, no sign-up, and they show their working
If you are not sure where to begin, start here — it sequences the material into a path rather than leaving you to guess an order.
How the material is written
Every article aims to do three things: state the formula, work through at least one example by hand, and explain where the concept actually shows up in machine learning. A derivative rule is more memorable when you can see it inside backpropagation.
The calculators are built to show intermediate steps rather than just an answer, because the steps are the part you are trying to learn. Worked examples are checked against those calculators before publication.
I am a student, not a professor. If you find an error — a wrong sign, a bad step, an explanation that does not hold up — I would genuinely like to know. Corrections improve the site and I would rather be right than look right.
Get in touch
Corrections, questions and suggestions for topics to cover are all welcome via the contact page.