Fundamentals of testing AI-based systems · Question 5 of 60

Why can a single test, such as a cat classified as a dog, not reliably reflect a model's functional correctness?

1 answer is correct. Easy

1. A Because AI systems work probabilistically and the same input does not always yield the same output
2. B Because the model is automatically retrained after every single test
3. C Because individual test cases are not permitted for AI systems under the EU AI Act
4. D Because cats and dogs are fundamentally indistinguishable for image classifiers