Question 13
Content Domain 3: ModelingA machine learning practitioner needs to choose compute resources for model training. Which choice directly reflects the resource categories identified for this decision?
Correct answer: B
Explanation
Selecting compute resources for model training involves choosing between processor types and execution patterns, such as GPU versus CPU and distributed versus non-distributed setups. — Choose appropriate compute resources (for example GPU or CPU, distributed or non-distributed).
Why each option is right or wrong
A. Select between batch and streaming data pipelines for training.
Compute resource selection is framed as GPU or CPU, and distributed or non-distributed.
B. Select between GPU or CPU resources and distributed or non-distributed training.
The source explicitly identifies compute resource selection for training as choosing, for example, GPU or CPU and distributed or non-distributed. This option matches both categories named in the material.
C. Select between supervised and unsupervised learning approaches.
Learning approach selection is separate from choosing GPU or CPU and distributed or non-distributed resources.
D. Select between training and inference evaluation metrics.
Metric selection is not the compute resource decision described; the named categories are processor type and distribution pattern.