Clustering
An unsupervised task that groups similar examples together based on distance or density.
- K-means: compact spherical groups around centroids.
- DBSCAN: dense regions of arbitrary shape.
- Hierarchical: tree of nested merges.
When to use: When you need to find natural groups in unlabeled data: segmentation, exploration, anomaly hunting.
Example: Group 50,000 e-commerce customers by recency, frequency, monetary (RFM). K-means with k=5 produces interpretable cohorts the marketing team can target.