You can build with AI tools without the math, but you can't debug or improve a model you don't understand. This module covers the math that shows up again and again in machine learning: linear algebra (vectors, matrices, determinants, eigenvalues) for representing data and transformations, and calculus (limits, derivatives, integrals) for measuring how a model's error changes as its parameters change.
It then covers statistics and probability: distributions, the normal curve, z-scores, sampling, and the central limit theorem, which underpin how you evaluate a model and judge whether a result is trustworthy. The final lesson ties it together with the math behind neural networks and backpropagation.
These courses are long, so treat them as a reference: watch the parts you need, then come back for the rest. When you're done, continue on to Data Handling & Analysis.
Vectors and their operations, the dot product, linear systems and matrices, Gaussian elimination, and reduced row echelon form.
Matrix operations, determinants and inverses, linear transformations, and eigenvalues and eigenvectors, including their role in dimensionality reduction and PCA.
Limits and slope, derivative rules including the chain rule, then integrals and the Fundamental Theorem of Calculus, with examples worked in Python and SymPy.
Sampling and experimental design, histograms and distributions, measures of center and spread, correlation, the normal distribution, z-scores, and the central limit theorem.
Vectors, matrices, determinants, limits, derivatives, and integrals, then how they connect to neural network architecture, activation functions, and backpropagation.