Unsupervised Learning
Sorting Strangers at a Party
6 min read
Sorting Strangers at a Party
You're at a party with 200 strangers and no nametags — but you start noticing patterns and naturally grouping people.
Without anyone telling you, you notice some people talk about cricket non-stop, others cluster around the food discussing recipes, and a third group is debating startup ideas. You've grouped them by behaviour without any labels. Unsupervised Learning does this with data — finds hidden structures and clusters without being told what to look for.
In Plain English
Unsupervised Learning finds hidden patterns in data without any labels or correct answers provided. The AI explores the data and discovers structure by itself.
The Technical Picture
Unsupervised learning algorithms (clustering, dimensionality reduction, generative modelling) find latent structure in unlabelled data. Common approaches include k-means clustering, DBSCAN, PCA, t-SNE, and autoencoders.
Real-World Examples
- Customer segmentation grouping buyers by behaviour without predefined categories
- Spotify discovering genre clusters from listening patterns
- Fraud detection finding unusual transaction clusters
Unsupervised Learning discovers hidden patterns in data — no labels, no answer key needed.