Skip to content
learn
  1. Learn
  2. Machine Learning
  3. Mathematics for ML (only what you need, visually)

ML-M01

Mathematics for ML (only what you need, visually)

Only the math machine learning actually uses, built visually: vectors, matrices, derivatives, probability, statistics, optimization and information theory.

10 lessonsabout 135 min

You do not need a math degree to do machine learning, but you do need a handful of ideas so solid that you can picture them. This module gives you exactly those: what a vector and a matrix really are, why the derivative is the engine of learning, how probability and statistics let you reason about noisy data, what training is optimizing, and why the classification loss is called cross-entropy.

Every lesson starts with a picture you can drag around, then shows the formula, then a worked example with real numbers, then a few lines of NumPy you can run. If you have forgotten school algebra, that is fine: everything is rebuilt from the beginning.

Lessons

  1. ML-01.01 Vectors and vector spaces A vector is a list of numbers, which you can picture as an arrow in space. Beginner 12 min
  2. ML-01.02 Matrices and linear transformations A matrix is a grid of numbers that transforms vectors: multiplying a vector by a matrix can rotate, stretch or project it. Beginner 14 min
  3. ML-01.03 Eigenvectors, eigenvalues and SVD Eigenvectors are directions that a matrix only stretches, never rotates, and eigenvalues say by how much. Intermediate 15 min
  4. ML-01.04 Derivatives and the chain rule A derivative measures how fast a function's output changes as its input changes, the slope at a point. Beginner 14 min
  5. ML-01.05 Probability fundamentals Probability measures how likely events are, from 0 to 1. Beginner 15 min
  6. ML-01.06 Distributions you will meet Distributions describe how likely each outcome is. Beginner 12 min
  7. ML-01.07 Statistics: sampling, CLT and confidence intervals We rarely see a whole population, only samples. Intermediate 15 min
  8. ML-01.08 Hypothesis testing and p-values A hypothesis test asks whether an observed effect could easily be chance. Intermediate 14 min
  9. ML-01.09 Optimization basics Training a model is optimization: finding parameters that minimize a loss function. Intermediate 12 min
  10. ML-01.10 Information theory for ML Information theory measures surprise. Intermediate 12 min

Start with lesson 1