ACorrect
Amazon SageMaker Ground Truth
Correct. Ground Truth is a data labeling (annotation) service that adds labels to training data using a combination of human effort and automation.
An ML team wants to reduce the effort of preparing training data and building models. Which SageMaker capabilities support the work of labeling training data and automatically building a model from data without writing code? (Choose TWO.)
1 / 1
Select all that apply
CorrectA, B
Explanation
A question about choosing TWO capabilities that support labeling and automatic model building.
- 1「labeling training data」Labeling = Ground Truth
- 2「automatically building a model from data」AutoML = Autopilot
BCorrect
Amazon SageMaker Autopilot
Correct. Autopilot is an AutoML capability that builds a model by automatically performing preprocessing, algorithm selection, training, and tuning just from the data you provide, without writing code.
CIncorrect
Amazon SageMaker Endpoint
An Endpoint is a mechanism that hosts a trained model for inference.
It is not a capability that supports labeling or automatic model building, so this is incorrect.
DIncorrect
Amazon SageMaker Model Monitor
Model Monitor is a capability that monitors a production model for drift.
It is not a capability that supports labeling or automatic model building, so this is incorrect.
EIncorrect
Amazon SageMaker Feature Store
Feature Store is a repository for storing, sharing, and reusing features.
It is not a capability that supports labeling or automatic model building, so this is incorrect.
Key Takeaway
The capabilities that support pre-training work are Ground Truth (adds labels to data) and Autopilot (AutoML that builds a model automatically without code). On the other hand, Endpoint (inference hosting), Model Monitor (monitoring after deployment), and Feature Store (storing features) are capabilities of other phases, not labeling or automatic building.