Generalization
The ability of a machine learning model to perform well on new, unseen data that was not part of its training set.
In plain language
How well an AI performs on new data it has not seen before. The ultimate goal is an AI trained on past customer data that accurately predicts future customer behaviour.
Why this matters
Generalisation is a core measure of AI quality and reliability. Poor generalisation (where a model performs well in testing but fails in production) exposes organisations to operational and reputational risk. Governance frameworks should include testing protocols that evaluate generalisation across diverse scenarios.
Relevance
ImplementationGeneralisation determines whether AI systems perform reliably in production, directly affecting the validity of model validation and operational risk management.
Related terms
Putting generalization 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.