An organization is establishing its AI model validation program. Which activity is MOST important to include in independent validation of a predictive model?
Select an answer to reveal the explanation.
Short Explanation
Here's the deal — you need to replicate the developer's analysis on independently drawn samples because that's how you verify the model's performance generalizes and wasn't inflated by the developer's chosen dataset. This directly tests robustness and reproducibility.
Full Explanation
The correct activity is replicating the model developer's analysis on independently drawn samples to verify that results are not specific to the developer's chosen dataset. This is one of the most important validation activities because it tests whether the model's reported performance generalizes and was not artificially inflated by the developer's choice of test data, directly testing model robustness and reproducibility. Confirming that the model documentation was prepared on time is a governance check. Reviewing the business case that justified the model's development is important but does not validate model performance. Ensuring that the model was built using the organization's approved technology stack is an IT governance concern.