K-Means

ELI5 - The Vibe Check

K-Means is a clustering algorithm that groups your data into a set number of buckets without ever being told what the buckets should mean. You say find me 5 groups, and it plants 5 random center points, assigns every data point to its nearest center, moves each center to the middle of the points assigned to it, and repeats until the centers stop moving. You get 5 clusters; figuring out what each one actually represents is on you.

Real Talk

K-Means is an unsupervised clustering algorithm that partitions data into k groups by iteratively assigning each point to its nearest centroid and recomputing centroids as the mean of their assigned points, until the assignments stabilize. It requires choosing k in advance, commonly guided by the elbow method or silhouette score, and works best when clusters are roughly spherical and similarly sized.

Show Me The Code

from sklearn.cluster import KMeans

kmeans = KMeans(n_clusters=5, random_state=42)
labels = kmeans.fit_predict(X)

When You'll Hear This

"We ran K-Means on purchase behavior and found 4 pretty distinct customer segments." / "K-Means assumed round, evenly-sized clusters, but our data's shaped nothing like that, results were garbage."