Interactive Notebook
Machine learning, made visible
Small, hands-on demos for building intuition about the linear algebra behind machine learning. Each one lets you poke at the inputs and watch the math respond.
Perceptron, step by step
Watch θ and its decision boundary update one point at a time — see how much the line jumps around before it settles into a clean separator.
GeometryMargins & distance
See how ‖θ‖ sets the width of the margin, project a point onto the boundary, and watch the distance d = |θ·x₀+θ₀|/‖θ‖ — with the full derivation underneath.
RegularizationAdjusting λ
Turn the regularization dial on the SVM objective `avg loss + (λ/2)‖θ‖²` and watch the boundary trade fit for a wider margin — with the optimizer running live.
ClusteringEM for Gaussian mixtures
Step through the E and M half-steps separately — watch points blend colors as responsibilities shift, the bells re-fit from soft counts, and ℓ climb (sometimes into a local optimum).
More demos
New visualizations drop in here as the set grows.