Machine learning Flashcards
(18 cards)
Which two AWS Machine Learning services are guaranteed to appear on the exam?
Amazon Rekognition and Amazon SageMaker.
What is Amazon Comprehend used for?
Sentiment analysis, indexing/searching reviews, legal brief management, and processing financial documents.
Which AWS service builds intelligent search over unstructured text?
Amazon Kendra.
What is Amazon Textract used for?
Converting text, handwriting, and data from scanned documents into digital text (OCR).
Which service is best for analyzing time-series data and making predictions?
Amazon Forecast.
When would you use Amazon Fraud Detector?
When building a customizable ML fraud detection model using your own data.
What does Amazon Transcribe do?
Converts speech (from video/audio) into text, e.g., for closed captioning.
What AWS service is used for building conversational chatbots?
Amazon Lex.
Which service converts text into lifelike speech in multiple languages?
Amazon Polly.
What is Amazon Rekognition used for?
Image and video analysis, including content moderation for social media.
What is the main purpose of Amazon SageMaker?
Building, training, and deploying ML models in the cloud.
What are SageMaker Notebooks?
Jupyter notebooks in the AWS environment used for ML model development.
What is SageMaker Inference used for?
Deploying trained models and generating predictions.
What is SageMaker Ground Truth used for?
Creating and managing training data labeling jobs using human input and active learning.
What is SageMaker Neo used for?
Optimizing ML models for specific hardware architectures (e.g., Nvidia, ARM, Intel).
What is Elastic Inference in SageMaker?
A feature that reduces inference cost by using CPU instead of GPU when possible.
How does Amazon Translate work?
Uses deep learning to translate text between languages.
What extra capabilities does SageMaker offer for production workloads?
Automatic scaling and high availability.
