AWS Machine Learning Engineer - Basics Quiz
← Back to Quiz Home
This quiz covers the foundational services for ML on AWS, including SageMaker, Textract, Rekognition, and cost optimization strategies.
What are the three main lifecycle stages managed by Amazon SageMaker?
SageMaker provides integrated tools for Notebooks (Build), managed Training jobs (Train), and hosting Endpoints (Deploy).
Which service effectively extracts text, handwriting, and tables from scanned documents?
Textract goes beyond simple OCR by understanding the structure of forms and tables.
How can you lower the cost of SageMaker training jobs by up to 90%?
Spot training uses spare EC2 capacity. SageMaker handles the interruption and resumption of checkpoints automatically.
Bedrock provides serverless access to LLMs via an API.
Which SageMaker feature helps you detect "Data Drift" (input distribution changes) in production models?
Model Monitor compares real-time production data against a baseline dataset (training data) to find anomalies.
What is the difference between Fine-Tuning and RAG?
RAG is preferred for keeping the model up-to-date with company knowledge.
Which service would you use to detect objects, faces, and unsafe content in images and videos?
Rekognition provides pre-trained computer vision models via an API.
What is a "SageMaker Endpoint"?
Endpoints provide a secure, scalable interface for applications to consume models.
How do you securely connect a SageMaker Notebook to a private database in your VPC?
Running notebooks in a VPC ensures traffic stays on the private network.
What is "SageMaker Studio"?
Studio provides a single web-based visual interface for all ML development steps.
Which input mode streams data from S3 to the training instance to start training faster (FIFO)?
Pipe Moode avoids downloading the entire dataset to disk before training starts, saving startup time and disk space.
What is "Amazon Transcribe"?
Transcribe handles audio ingestion and generates transcripts with timestamps.
What is the primary benefit of "Multi-Model Endpoints" (MME)?
MME is ideal for SaaS applications where each customer has a custom fine-tuned model that is rarely accessed.
Which service converts text into lifelike speech?
Polly uses deep learning to synthesize natural-sounding human speech.
What is "Amazon Q" for AWS?
Amazon Q helps developers and admins work faster by answering technical questions in the console/IDE.
When should you use Batch Transform instead of an Endpoint?
Batch Transform spins up a cluster, processes the S3 data, and shuts down, saving money compared to a 24/7 endpoint.
How does SageMaker handle the underlying infrastructure for training?
SageMaker abstracts the heavy lifting of infrastructure management for training jobs.
What is the "Ground Truth" in the context of Model Monitor?
Without ground truth (feedback loop), you can detect data drift but not accuracy drift.
Which instance family is optimized for Deep Learning Training?
Training requires massive parallel processing power found in GPUs or Trainium chips.
What is "Local Mode" in the SageMaker SDK?
Local mode saves time and money by avoiding the spin-up overhead of a full training job during creating the script.
Quiz Progress
0 / 0 questions answered
(0%)
0 correct
Quiz Complete!
0%
📚 Study Guides
📬 Weekly DevOps, Cloud & Gen AI quizzes & guides