Question 11
Domain 3: Model DevelopmentA data scientist is tuning hyperparameters for a model that runs on a single machine, but the search is taking too long when evaluated sequentially. What is the most appropriate way to reduce total tuning time based on the stated guidance?
Correct answer: B
Explanation
When hyperparameter tuning is performed for a model running on a single node, parallel processing is used to speed up evaluation of tuning runs. — official.txt
Why each option is right or wrong
A. Increase the number of hyperparameters searched while keeping all trials sequential
Parallel processing changes execution speed; adding more hyperparameters usually increases tuning work.
B. Use parallel processing to evaluate hyperparameter tuning runs on the single node faster
The source states that parallel processing is used to speed up hyperparameter tuning for models running on a single node. Because the model already runs on one machine and the issue is sequential tuning time, parallelizing the tuning work is the supported approach.
C. Move the model to a different algorithm before attempting any hyperparameter search
Algorithm changes are not identified as the method for speeding hyperparameter tuning on a single node.
D. Disable hyperparameter tuning and rely on default model settings for faster iteration
The guidance addresses speeding tuning with parallel processing, not eliminating tuning altogether.