Data augmentation

Data augmentation is a technique that expands a training dataset by creating modified copies of existing examples, such as rotated images or paraphrased sentences, to improve robustness and reduce overfitting.

Data augmentation sits in the Foundations part of the Agentik {OS} glossary, which defines the words used to build and run AI agent systems.