ML technique used primarily in unsupervised learning that improves model performance by teaching the model to distinguish between similar and dissimilar data points.
Generality: 720