Welcome to Cluster Explorer! 🎉 K-Means finds hidden groups in data — without any labels! First, let's create a dataset. Click or drag on the canvas to drop data points. Or pick a preset dataset. Can you create 3 clear groups?
Step 1 — Create Your Dataset
0 points
Presets:
💡 Click anywhere to add a point · Click & drag to paint multiple
📊 Dataset Info
Total points0
Canvas area480 × 400
6
🔵
Wizzy the AI Tutor
Now choose K — the number of clusters! Look at your data and guess how many groups there are. Then you can click on the canvas to manually place your starting centroids, or let the algorithm pick them randomly!
Step 2 — Choose K & Place Centroids
💡 Click to place K centroid(s) — or use Random Init
🎯 Centroid Setup
K (clusters)3
Placed0 / 3
🔵
Wizzy the AI Tutor
Watch the magic! ✨ Click "Step" to see each assignment and centroid-move separately, or "Auto Run" to watch it converge automatically. The ★ stars are centroids — watch them drift toward the centre of their clusters!
Step 3 — K-Means in Action
0
Iterations
—
Current Phase
Ready
Status
—
Inertia (WCSS)
Press Step to begin the first assignment phase, or Auto Run to watch the full algorithm!
📊 Clustering Stats
Iteration0
Total points—
Inertia (WCSS)—
Centroid shift—
Converged?No
3
📉 Inertia over Iterations
🔵
Wizzy the AI Tutor
But how do you choose the right K? The Elbow Method runs K-Means for K=1 to 8 and plots the inertia. The "elbow" bend is where adding more clusters stops helping much. Can you spot the elbow in your data?
Step 4 — The Elbow Method
🔍 Elbow Analysis
Best K (suggested)—
Your chosen K—
Data points—
What is Inertia?
Inertia = sum of squared distances from each point to its cluster centroid. Lower = tighter clusters. Adding more clusters always reduces inertia, but the rate of decrease slows after the "elbow" — that's the natural K!
🔵
Wizzy the AI Tutor
🎊 You've just run the same algorithm used in Spotify playlist grouping, customer segmentation, image compression, and medical diagnosis! K-Means is one of the most powerful unsupervised learning tools. You are now a Clustering Scientist! 🌟
🔵
Clustering Scientist Badge!
You ran K-Means clustering and used the Elbow Method. Enter your name for your certificate!
Optional. Stays on this device only — not sent to WhizzStep.
🔵 WhizzStep AI Lab
Activity completion card for
Student Name
has explored K-Means Clustering & Unsupervised Learning
This records completion of a browser activity only. It is not an accredited certificate or proof of mastery.
K-Means Master
Elbow Expert
Cluster Scientist
Clustered — points · K=— · whizzstep.in
Key Concepts You've Mastered
Unsupervised Learning
🔓 No Labels Needed
Unlike classification, clustering finds patterns in data without any predefined categories or labels.
Centroid
⭐ The Cluster Centre
A centroid is the mathematical average position of all points in a cluster. It moves each iteration.
Inertia (WCSS)
📏 Cluster Tightness
Within-Cluster Sum of Squares. Measures how spread out points are within each cluster.
Elbow Method
💪 Choosing K
Plot inertia for K=1..8. The "elbow" bend shows where adding more clusters gives diminishing returns.
Convergence
🛑 When It Stops
K-Means converges when centroids stop moving — the clusters have stabilised into a final answer.
Real World Uses
🌍 Where It's Used
Spotify playlists, customer groups, image compression, medical diagnosis, news grouping, and more!
About this lab
AI-11How AI and Machine Learning WorkClasses 8-1015 minApplied
Learning objective: Group unlabelled points into clusters and see how an unsupervised algorithm finds structure without being told the answer.
What this simplifies: A small 2D dataset is used for visual clarity; real clustering works across many more dimensions.
Privacy: No learner input leaves the device.
Teacher prompt: Ask the class why this simulation might mislead someone who takes it too literally.
Reflect: What is one thing this activity showed you that you did not expect?
