Question 1
Domain 3: Application DevelopmentA team trains several candidate models for the same task and wants to choose the one to deploy. Based on the source material, what is the most appropriate basis for selecting the best model?
Correct answer: B
Explanation
Model selection for a task is done by comparing common metrics generated during experiments and choosing the model with the best task performance. — official.txt
Why each option is right or wrong
A. Choose the model that was trained first, because earlier experiments establish the baseline choice.
Selection is based on comparing common metrics from experiments, not training order.
B. Choose the model that shows the strongest task performance when comparing common experiment metrics.
The source material states that selecting the best model for a task is done by comparing common metrics produced during experiments. Because the goal is model selection for task performance, the correct choice is the model with the strongest performance on those common metrics.
C. Choose the model with the largest number of experiments, because more runs automatically indicate better quality.
The criterion is task performance on common metrics, not the count of experiments.
D. Choose the model that uses the newest architecture, because newer designs outperform older ones by default.
Model choice depends on experimental metrics for the task, not whether an architecture is newer.