Maths for ML
My notes on the maths behind machine learning, as seven visual, interactive guides. Each one starts from zero, explains things in plain English with worked examples, and has interactive visuals you can play with. Your chapter progress is saved in your browser.
Maths foundations
Read these in order, since each one builds on the one before:
- Linear Algebra for Machine Learning: vectors, matrices and the tools that models are built from.
- Calculus for Machine Learning: derivatives, gradients and how a model learns from its errors.
- Optimization for Machine Learning: gradient descent and its relatives, which put the first two to work.
Statistics for ML
A four-guide series, also best read in order:
- Probability and Data: probability, random variables, distributions, descriptive statistics, correlation, KDE and Q-Q plots.
- Estimation, Inference and Experiments: MLE and MAP, hypothesis tests, confidence intervals, A/B testing, regression, GLMs, PCA and t-SNE.
- Bayesian Modeling and Computation: priors, conjugacy, hierarchical models, MCMC, HMC and NUTS, variational inference.
- Time Series and Bayesian Forecasting: autocorrelation, stationarity, classical forecasting, a Prophet-style Bayesian model and forecast evaluation.