Unsupervised Learning
Find structure in data that has no labels, using k-means clustering and PCA for dimensionality reduction.
No labels, no right answer
In supervised learning each example comes with a label to predict. In unsupervised learning you only have inputs, and you ask the data to show its own structure.
| Task | Question | Common method |
|---|---|---|
| Clustering | Which examples belong together? | k-means |
| Dimensionality reduction | Can fewer numbers describe each example? | PCA (principal component analysis) |
| Anomaly detection | Which examples look unlike the rest? | Distance or density based |
Typical uses: group customers into segments, compress features before another model, plot high-dimensional data in 2-D, find unusual transactions.
1 of 5