Machine learning math

These 67 lessons cover the mathematical ideas used to describe data, train models, and evaluate results. The seven modules span statistics, probability, linear algebra, calculus, optimization, graph theory, and information theory. Choose a module from the sidebar, or open the full lesson index below to find a specific concept.

If you are starting out, begin with Basic Statistics to review averages and variance. Continue with Conditional Probability, then Dot Product & Vector Norms. These topics provide a useful foundation before working through derivatives and model training.

For the math behind training, follow Partial Derivatives with Backpropagation & Gradient Descent, then explore optimization methods. You can study graph theory and information theory whenever they become relevant to a model or problem you are working on. Each module lists its lessons, so you can follow this path or jump directly to the concept you need.