Robustness and stress testing
Checking that a model holds up under unusual, noisy or extreme inputs.
Tests include edge cases, missing or corrupted values, distribution shifts and deliberately perturbed inputs. A robust model degrades gracefully instead of failing silently.
On the exam
Testing only on clean data that looks like the training data doesn’t show how the model behaves in the real world.
Related terms
- Model validation and testingChecking that a model is fit for its intended use before and after deployment.
- Adversarial examples and evasion attacksInputs crafted to fool a model.
- AI red teamingDeliberately attacking an AI system to find weaknesses before others do.
AAIA Test yourself on this with timed AAIA practice questions in the AAIA Exam Simulation Hub, or read the free AAIA exam guide.
Test yourself with today’s question of the day.
