Singular Value Decomposition (SVD)

A mathematical technique that breaks down a complex matrix into simpler component matrices. In machine learning, it's used for tasks like dimensionality reduction and recommender systems by identifying underlying patterns and noise in data.