Data Augmentation
Techniques for artificially expanding a training dataset by creating modified versions of existing data, such as rotations, translations or paraphrasing, to improve model robustness and generalisation.
In plain language
Creating more training data by modifying existing data. Flipping images, adding noise, paraphrasing text; it's like multiplying your dataset to make the AI work better without collecting new data.
Why this matters
Data augmentation directly improves AI model robustness and reduces overfitting, which affects reliability in production. Your AI governance framework should require documentation of augmentation techniques used and their impact on fairness and bias across different demographic groups.
Relevance
ImplementationData augmentation is a technical control that improves model performance and robustness, essential during AI system development and testing.
Related terms
Putting data augmentation into practice in your organisation?
Ready to transform your AI strategy?
Partner with Australia's AI strategy and governance specialists. From adoption roadmaps to ISO 42001 audit readiness.