ML model risks · Question 21 of 60
Which types of risk does ML model testing address?
3 answers are correct. Medium
Show answer
Answer
- B Functional risks such as bias, overfitting and adversarial vulnerabilities
- D Non-functional risks such as lack of robustness and performance efficiency
- E Deployment risks
Explanation
ML model testing deals with functional risks such as bias, overfitting and susceptibility to deception, with non-functional risks such as lack of robustness or predictions that are too slow, and with risks during rollout.