ML Math Lab — K-Means: Star Colony
Stars drop → particles get claimed → stars migrate → colonies stabilize (GOD MODE)
Phase: Idle
Hint: press Start
Iterations: 0
Section 1 — What you will learn
✓ By the end you will understand
  • K: how many colonies (groups) the AI tries to form.
  • Assignment step: each point joins its nearest star (centroid).
  • Update step: each star moves to the average of its claimed points.
  • Convergence: stars stop moving → the solution stabilizes.
  • Distance metric: Euclidean vs Manhattan changes “nearest”.
  • Elbow method: how to pick a good K using inertia (total mess energy).
How to play: Press Start to watch the cinematic 4 phases.
Click in the arena to add new particles (AI will re-balance).
Section 2 — Controls
Story helps kids remember what “grouping without labels” means.
Tip: try K=2 vs K=6 on the same dataset.
Noise is visual: wind lines, point jitter, and outlier meteors.

Section 3 — Live stats
Inertia (Total Mess Energy)
Centroid Movement
Convergence
Outliers (meteors)
Iterations
0
Best K (Elbow guess)
Live Chart
Inertia drops as AI improves
Elbow Method
Elbow = point after which adding more colonies barely reduces mess.

Section 4 — Live math (simple)
Loading…
🧠 Kid Coach
Loading…
Tip: Watch lasers (assignment) and star movement (update).