Accuracy is unestablished for affected seasonal products: What does this evidence most directly support?
Strong labeled accuracy cannot cover categories with systematically missing outcomes; their performance remains unestablished.
The question
A retail demand model was evaluated on a carefully sampled test set, but labels are missing for many products and concentrated in seasonal categories. The labeled subset shows strong accuracy. What does this evidence most directly support?
Preparing for AIGP? Take the free 5-min readiness quiz →
- The strong labeled accuracy represents all products.Selection and missingness concentrated in seasonal categories make broad generalization unsupported without evidence for those products.
- The unlabeled products can be treated as correct predictions.Unlabeled outcomes provide no correctness evidence and cannot be converted into validation results merely because predictions exist.
- Accuracy is unestablished for affected seasonal products. ✓Concentrated missing labels limit evaluation coverage, so strong results on labeled items cannot establish performance for affected seasonal products.
- Adding more labeled products automatically resolves the limitation.Additional labels help only if they cover the affected categories appropriately and are measured using suitable evaluation procedures.
The trap
Ask whether missing outcomes are concentrated; if so, labeled accuracy may not generalize. How to remember it
Strong labeled accuracy cannot cover categories with systematically missing outcomes; their performance remains unestablished.
How many of these would you get right?
One of 1581 AIGP questions on Certsqill. Take a free five-minute check and see your score per domain — not one number, but which section to open tonight.
Test your AIGP readiness — freeMore Understanding How to Govern AI Development questions
- Urban-rural performance is comparable in this test set: What does the evidence most directly show? →
- Historical labels may reproduce institutional bias: What does the evidence most directly indicate? →
- Postal-sector codes warrant proxy-bias analysis: What does this evidence most directly suggest? →
- All 426 Understanding How to Govern AI Development questions →
Part of the Certsqill AIGP question bank · Understanding How to Govern AI Development ·
Every answer, right and wrong, comes with its own explanation.