Brainscape's Knowledge GenomeTM


Top Flashcards (Certified)
All Flashcards

AWS Machine Learning Specialty > High Level ML Services > Flashcards

High Level ML Services Flashcards

1
Q

What is Amazon Comprehend

A

It is a managed ML Service focused on NLP.

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
2
Q

What are some examples of Amazon Comprehend capabilities?

A

-Sentiment Detection
-Entity Identification
-Text Analysis
-PII detection and redaction
-Key Phrase Extraction
-Language Identification
-Syntactical Analysis

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
3
Q

True or False: Amazon Comprehend can be trained on additional data

A

True

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
4
Q

What is Amazon Comprehend Medical and how is it different from regular Comprehend?

A

It is a comprehend tuned to medical data and documents, being more performant on them and being able to detect Protected Health Information (PHI) (DetectPHI API)

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
5
Q

What is Amazon Translate?

A

It is a managed ML Service focused text translation from one language to another.

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
6
Q

True or False: Amazon Translate accepts inputs in CSV and TMX Format

A

True

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
7
Q

What is the main use of AWS Transcribe?

A

Converting audio to text

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
8
Q

How can Transcribe deal with confidential information on the audio being transcribed?

A

Transcribe can automatically remove Personally Identified Information using Redaction, but any other kind of confidential information will still pass through

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
9
Q

Declare True or False for the statements below:
-Transcribe accepts only MP3 and MP4 as inputs, aside from streaming audio
-Transcribe is incapable of differentiating different speakers on the same audio
-Transcribe automatically identifies the language being spoken

A

-False, it accepts FLAC and WAV as well
-False, it can differentiate speakers as long as you provide the number of speakers in the audio
-True
-Transcribe supports the usage of custom vocabularies to increase transcription quality

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
10
Q

What languages does Transcribe Streaming accept?

A

English, Spanish and French

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
11
Q

What is the main use of AWS Polly?

A

Converting text to audio

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
12
Q

True or False: Polly is incapable of customizing the pronunciation of words

A

False, this customization can be done through Lexicons in the SynthetizeSpeech operation

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
13
Q

True or False: Polly pronunciation can be manipulated through the use of SSML on the input text

A

True

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
14
Q

True or False: Speech marks can be used to encode where a sentence begins and ends on Polly

A

True

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
15
Q

What are the main uses of AWS Rekognition?

A

Locating persons, objects, texts and scenes in images or videos.

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
16
Q

Rekognition can use facial analysis and facial search to do people counting and user verification. What are the 2 types of face Rekognition can detect?

A

Rekognition can detect either a specific face from a catalog of faces or compare faces to famous celebrities.

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
17
Q

How could AWS Rekognition be used to help with content moderation?

A

It can be used to detect inapropriate images or videos. Those can then be removed or flagged for manual review.

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
18
Q

True or false: The Confidence Threshold for items to be flagged on moderation cases for Rekognition can be manually configured

A

True

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
19
Q

True or False: When working with video, Rekognition inference data must come from S3, and prioritize higher framework over resolution

A

False, data must come from Kinesis Video Streams, and should prioritize higher resolution

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
20
Q

How would you use Rekognition to detect labels it wasn’t trained on?

A

Retrain the model using images with custom labels

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
21
Q

What is the main use of Amazon Forecast?

A

ML service that delivers forecasts based on time-series data

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
22
Q

True or False: The AutoML Forecast feature allows Forecast to select the best model for your time series data

A

True

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
23
Q

List all of Amazon Forecast’s algorithms

A

-CNN-QR
-DeepAR+
-Prophet
-NPTS
-Arima
-ETS

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
24
Q

Describe each of the following Amazon Forecast algorithms:
-CNN-QR
-DeepAR+
-Prophet
-NPTS
-Arima
-ETS

A

-CNN-QR: Convolutional Neural Network – Quantile Regression. Best for datasets with hundreds of time series, accepts related historical time series data & metadata
-DeepAR+: RNN. Best for large datasets, accepts related forward-looking time series & metadata
-Prophet: Additive model with non-linear trends and seasonality
-NPTS: Non-Parametric Time Series. Good for sparse data. Has variants for seasonal / climatological forecasts.
-ARIMA: Autoregressive Integrated Moving Average. Commonly used for simple datasets (<100 time series).
-ETS: Exponential Smoothing. Commonly used for simple datasets (<100 time series).

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
25
What is the main use of AWS Lex?
Creating chatbots, specially voice operated ones
26
What is the main use of Amazon Connect?
To setup a virtual call center
27
How do Amazon Lex intents work?
The inputs received by Lex can invoke intents, which then invoke lambda functions to fulfill them.
28
What is Amazon Lex Automated Chatbot Designer?
It is a feature from Lex that creates a Chatbot automatically based only on a list of existing conversation transcripts
29
What is Amazon Personalize?
It is a fully managed recommendation engine
30
True or False: API Access to Amazon Personalize requires the use of Javascript or the SDK
True
31
What kinds of Recommendation API Endpoints does Personalize have?
-GetRecommendation: Recommeds similar products -GetPersonalizedRanking: Ranks a list or items provided, allowing editorial control/curation
32
True or False: Amazon Personalize can only perform real-time recommendations
False, it can also perform batch recommendations
33
What are Amazon Personalize contextual recommendations?
They are recommendations performed based on contextual information, such as device type, time, etc.
34
Explain each of the following Amazon Personalize features: - Business rules and filters - Promotions - Trending Now - Personalized Ranking
-Business rules and filters: Personalize allows you to filter out recently purchased items, highlight premium content or make sure a percentage of recommendations are of a certain category -Promotions: Personalize allows recommendations to be influenced by current promotions -Trending Now: Personalize can recommend the most popular itens at the moment -Personalized Ranking: Personalize can personalize its results, promotions and curated lists based on the client
35
Describe the following Amazon Personalize terminology: -Datasets -Recipes -Solutions -Campaigns
-Datasets: Users, items and interactions used to train the models -Recipes: What a model should be trained to do (PERSONALIZED_RANKING, RELATED_ITEMS, etc) -Solutions: A trained version of the model based on a dataset and recipe -Campaign: Deployment of solution version
36
How can one keep Personalize models current?
-Keep datasets updated -Feed user interactions to the model using the PutEvents API -Retrain the model partially every 2 hours and fully every week
37
How does Amazon Personalize pricing work?
- Data ingestion: per-GB - Training: per training-hour - Inference: per TPS-hour - Batch recommendations: per user or per item
38
What are Amazon Transcribe Personalized use cases?
-Call Analytics -Medical -Subtitling
39
True or False: You can use Amazon Lambda to trigger Rekognition processing of an image whenever it is saved on S3
True
40
What is the main use of Amazon Textract?
Convert image text and data into computer format
41
What is Amazon DeepRacer?
It is a reinforcement learning powered racing car
42
What is Amazon Lookout?
It is a service that analyzes sensor data to detect abnormalities indicating equipment issues
43
What is Amazon Monitron
End to end system for monitoring industrial equipment & predictive maintenance
44
What is TorchServe?
A model serving framework for Pytorch
45
What is AWS Neuron?
It is a SDK for optimizing inference using AWS Inferentia chips that can be integrated withn Sagemaker and others AMIs and frameworks
46
What is AWS Panorama?
It is a service for executing computer vision at the edge on existing IP cameras
47
True or False: TorchServe is a joint of work between Amazon and Meta on the PyTorch project
True
48
True or False: To use AWS Neuron on Sagemaker, you need to customize your inference image to use it
False, Sagemaker implements it by default
49
What is AWS DeepComposer?
It is an AI powered keyboard that composes song from a melody. It is only for educational purposes.
50
What is Amazon Fraud Detector?
It is a service that allows you to build your own custom fraud detector model from a template you choose
51
What is Amazon CodeGuru?
It's a service that uses ML to automates code reviews by identifying code maintainability and security issues and perform application recommendations.
52
What are CodeGuru's main functionalities?
-CodeGuru Reviewer: Automates code review for static code analysis (development) -CodeGuru Profiler: Visibility/recommendations about application performance during runtime (production)
53
True or False: Amazon CodeGuru Reviewer supports Java, JS, Python and Ruby
False, only Python and Java
54
What is Contact Lens for Amazon Connect?
It is a feature that facilitates call center analysis, ingesting audio data from the calls and analyzing it from that.
55
What are some of Amazon Contact Lens features?
-Allows search on calls and chats -Can perform sentiment analysis -Find "utterances" that correlate with successful calls -Categorize calls automatically -Measure talk speed and interruptions -Theme detection: Discovers Emerging issues
56
What is Amazon Kendra?
It an enterprise search service with natural language.
57
Where can Amazon Kendra data be stored?
-Sharepoint -Intranet -Sharing Services (JDBC, S3)
58
What is Amazon Augmented AI (AI21)?
It is a service that standardizes human review of low-confidence ML predictions
59
What are the Amazon Integration with Amazon Augmented AI (AI21)?
-Mechanical Turk -Textract -Rekognition -Sagemaker
60
What are the model Recipes available for Amazon Personalize?
-USER_PERSONALIZATION -PERSONALIZED_RANKING -RELATED_ITEMS
61
What are the hyperparameters for Amazon Personalize "User-personalization" and "Personalized Ranking" recipes?
-hidden_dimension (HPO) -bptt (backpropagation through time - RNN, true or false) -recency_mask (weight recent events) -min/max_user_history_length_percentile (filters out robots) -exploration_weight (0-1) -exploration_item_age_cutoff (explores how far back in time you go)
62
What are the hyperparameters for Amazon Personalize "Similar-items" recipes?
-item_id_hidden_dimnesion (HPO) -item_metadata_hidden_dimension (HPO with min and max range specified)
63
What are the different versions of Amazon Lookout?
-Lookout for metrics -Lookout for equipment -Lookout for vision
64
How does lookout for vision detect anomalies?
Through the usage of computer vision algorithms and camera feeds

AWS Machine Learning Specialty (7 decks)

Brainscape helps you reach your goals faster, through stronger study habits.
© 2025 Bold Learning Solutions. Terms and Conditions