Data scientists need to prepare data, train a model, and deploy it on AWS.
Choose a managed ML development platform.
Which service is most appropriate?
Amazon SageMaker AI supports building, training, deploying, and monitoring machine-learning models.
Detailed explanation
S3 can store data, but it does not by itself manage training or model endpoints.
S3 can store data, but it does not by itself manage training or model endpoints.
It supports the machine-learning model lifecycle from development and training through deployment and monitoring.
It supports the machine-learning model lifecycle from development and training through deployment and monitoring.
Budgets monitors cost or usage thresholds rather than developing models.
Budgets monitors cost or usage thresholds rather than developing models.
Route 53 provides DNS and domain services, not an ML development platform.
Route 53 provides DNS and domain services, not an ML development platform.
Check official documentation
Compare the following points with the official documentation.
Amazon SageMaker AIの公式概要で、訓練・デプロイ・推論の機能を確認する。
Official references
What to confirm
モデル開発・訓練・デプロイの要件へSageMaker AIを対応付けられる。
Key points
- SageMaker AI
- Model training
- Model deployment
Notes
- Environment:
- AI/ML and analytics service features can change; confirm current behavior in official AWS documentation.
Foundation review
Input and output direction
Separate image analysis, speech-to-text, and text-to-speech by the direction and type of transformation.
Pipeline stage
A query, model inference, dashboard, and data-ingestion service solve different stages of an analytics workflow.