Machine Learning

K-Means Clustering Step by Step

Click to place K centroids on the canvas, then step through iterations watching assignment and centroid update alternate until convergence.

Iterations run: 0/5

Observed bad init

Place 3 centroids (click canvas)

⭐ = centroid

Algorithm Setup

Number of clusters (K)3

23456

What's Happening

Click on the scatter plot to place 3 starting centroids. These are the initial cluster centers.

Elbow Curve

Run the algorithm with different K values to build the elbow curve

Run with K=2..6 to build the curve. The "elbow" suggests the optimal number of clusters.