Data pipeline and representativeness · Question 34 of 60
How can the expected operational data distributions and critical edge cases be identified when defining the target population?
3 answers are correct. Medium
Show answer
Answer
- A By using benchmark datasets from trusted sources such as NIST
- C By analyzing data from existing or similar systems
- D By consulting domain experts on real-world data patterns
Explanation
The expected distributions and critical edge cases are identified by consulting domain experts, analyzing data from existing systems or similar applications and using benchmark datasets from trusted sources such as NIST or industry databases. Stratified sampling is then used to create a baseline that covers all relevant subgroups.