Distance metrics (Euclidean, Manhattan, Minkowski) define how 'close' two data points are, which is the foundation for algorithms like KNN and K-Means.
The geometry of learning.
Critical choice in clustering and vector search.