◇ CLUSTER

Watch a machine find the groups.

Scatter points on a table — you see the clumps at a glance; a machine starts blind. k-means plants a few guesses, then repeats two moves: assign each point to its nearest guess, move each guess to the middle of what claimed it. Watch the guesses migrate and the groups fall out of the noise. No labels — it finds the structure itself.

blobs stretched moons rings
Hand-written k-means · numpy · $0. The third way a machine learns, with forage (RL) & descent (supervised). Part of the Stateful AI galaxy.

cluster — finding the groups

blobsstretched moonsrings
ITERATION
INERTIA
SEPARATION