Explanation:
The correct answers are C and A. In a regulated environment, traceability and explainability are crucial for compliance and understanding model decisions, while reproducibility ensures consistent results. Federated learning (A) is also important for privacy-preserving data analysis, especially in sensitive sectors like insurance. However, the primary focus should be on factors that directly impact regulatory compliance and ethical considerations, making C the best single answer when not choosing two. For more details, refer to the OECD document.
Ultimate access to all questions.
No comments yet.
As an ML engineer at a regulated insurance company, you're tasked with developing a model to approve or reject insurance applications. The model must comply with strict regulatory requirements, ensure fairness, and maintain high accuracy. Given these constraints, which of the following sets of factors should you prioritize to meet both regulatory and ethical standards? (Choose two options)
A
Federated learning, reproducibility, and explainability
B
Differential privacy, federated learning, and explainability
C
Traceability, reproducibility, and explainability
D
Redaction, reproducibility, and explainability
E
All of the above