Back to Glossary
Unsupervised Learning
Fundamentals
Finding structure in unlabeled data (e.g., clustering).
Unsupervised learning discovers patterns without predefined labels.
- Types: Clustering, dimensionality reduction, density estimation, anomaly detection.
- Applications: Exploratory analysis, preprocessing, recommender pretraining.
- Risks: Subjective interpretation, sensitivity to parameters and distance metrics.
- Practice: Visualization, stability checks, incorporation of domain knowledge.