ML model risks · Question 42 of 60
Which mappings of model risk and mitigation are correct?
3 answers are correct. Hard
Show answer
Answer
- C Failure to achieve performance measures such as accuracy or recall: ML functional performance testing
- D Unethical model: ethical system testing
- E Functional incorrectness and non-functional defects: metamorphic testing
Explanation
Ethical system testing addresses an unethical model, metamorphic testing addresses functional incorrectness and non-functional defects, and ML functional performance testing addresses failure to achieve performance measures. Overfitting is addressed by testing for overfitting, and an API defect by API testing.