Cheat sheets
The formulas, definitions and classic mistakes for each part of a machine learning or AI course, on two printed pages. Every idea links to a lesson or lab where you can watch it work.
Supervised learning cheat sheet
Linear models and losses, Gradient descent, kNN and naive Bayes, Trees and SVMs
20 ideas, 6 classic mistakes
Model evaluation cheat sheet
Confusion matrix metrics, Curves, regression metrics and calibration, Validation, Bias, variance and regularisation
18 ideas, 6 classic mistakes
Neural networks cheat sheet
Neurons and layers, Activations, Backpropagation, Training, Convolutional networks
19 ideas, 6 classic mistakes
Unsupervised learning cheat sheet
k-means, Other clustering, Principal component analysis, t-SNE and UMAP
18 ideas, 6 classic mistakes
NLP and transformers cheat sheet
Tokens and embeddings, Recurrent networks, Attention, The transformer block, Counting parameters
18 ideas, 6 classic mistakes
Reinforcement learning and search cheat sheet
MDPs and returns, Bellman equations and planning, Learning from experience, Search
18 ideas, 6 classic mistakes
Probability and maths for ML cheat sheet
Probability rules, Expectation, variance and distributions, Estimation and information, Linear algebra, Derivatives you need
18 ideas, 6 classic mistakes
Large language models cheat sheet
Objective and evaluation, Sampling, Scaling, Post-training, Systems
18 ideas, 6 classic mistakes
Following a course? The course companions point to the right sheet each week.