1
Q
What does Artificial Intelligence (AI) refer to?
A
The simulation of human intelligence in machines designed to think and act like humans.
2
Q
Name three applications of AI.
A
- Personal Assistants
- Fraud Detection
- Medical Imaging
3
Q
What is Machine Learning (ML)?
A
A method of data analysis that automates analytical model building.
4
Q
What are the types of data in Machine Learning?
A
- Structured Data
- Semi-Structured Data
- Unstructured Data
5
Q
Fill in the blank: In ML, the _______ process involves feeding large amounts of data to an algorithm.
A
training
6
Q
What are the three styles of Machine Learning?
A
- Supervised Learning
- Unsupervised Learning
- Reinforcement Learning
7
Q
What is Deep Learning?
A
A subset of ML that utilizes neural networks with multiple layers to recognize complex patterns.
8
Q
What is the structure of a neural network?
A
- Input Layer
- Hidden Layers
- Output Layer
9
Q
Name two applications of Deep Learning.
A
- Image Classification
- Natural Language Processing (NLP)
10
Q
True or False: Traditional ML requires manual feature selection.
A
True
11
Q
What is Generative AI?
A
Models that generate new content based on training data.
12
Q
What are transformers in Generative AI?
A
A model architecture that processes sequences of data in parallel.
13
Q
Fill in the blank: In Generative AI, _______ encodes the relative position of each token in a sequence.
A
Positional Encoding
14
Q
What is the difference between few-shot and zero-shot learning?
A
- Few-shot Learning: Provides a few examples
- Zero-shot Learning: Performs a task without examples
15
Q
What are the advantages of using AWS Generative AI Services?
A
- Accessibility
- Efficiency
- Cost-Effectiveness
- Speed to Market
- Security and Compliance
- Meeting Business Objectives
16
Q
What are some limitations of Generative AI?
A
- Lack of Contextual Understanding
- Hallucinations and Inaccuracy
- Interpretability Challenges
17
Q
What characterizes Supervised Learning?
A
The model is trained on a labeled dataset to predict outputs for new data.
18
Q
What are the two types of Supervised Learning?
A
- Classification
- Regression
19
Q
What is Clustering in Unsupervised Learning?
A
Grouping data points into clusters based on similarity.
20
Q
What is Semi-Supervised Learning?
A
A blend of supervised and unsupervised learning using both labeled and unlabeled data.
21
Q
What is Reinforcement Learning?
A
An agent learns to make decisions by maximizing cumulative rewards through actions.
22
Q
What is the difference between Batch Inference and Real-Time Inference?
A
- Batch Inference: Delayed results for large numbers of inferences
- Real-Time Inference: Immediate responses to client requests
23
Q
What is a confusion matrix?
A
Summarizes classification model performance with true positives, true negatives, false positives, and false negatives.
24
Q
What is Precision in model evaluation?
A
Focuses on the accuracy of positive predictions.
25
What are the regression metrics used to evaluate models?
* Mean Squared Error (MSE)
* Root Mean Squared Error (RMSE)
* Mean Absolute Error (MAE)
26
What does Amazon Rekognition service do?
Deep learning service for computer vision tasks.
27
What is Amazon Comprehend used for?
NLP service that discovers insights and relationships in text.
28
True or False: AWS provides pre-trained AI services accessible through APIs.
True
29
What is the function of transcription in AI?
Converts spoken language into text (speech-to-text)
## Footnote
Includes real-time transcription and captioning for live or recorded audio/video.
30
What does Amazon Personalize provide?
Provides personalized recommendations for customers
## Footnote
Includes product/content recommendations and targeted marketing campaigns.
31
What is the purpose of Amazon Translate?
Neural machine translation for text across 75 languages
## Footnote
Supports real-time translation in chat applications and multilingual content creation.
32
What is Amazon Forecast used for?
AI service for time series forecasting
## Footnote
Includes demand forecasting, inventory management, and financial planning.
33
What does Amazon Fraud Detector do?
Detects potentially fraudulent online activities using pre-trained models
## Footnote
Helps prevent online payment fraud, detect fake accounts, and prevent account takeover.
34
What is Amazon Bedrock?
Service to build generative AI applications using foundation models from top AI providers
## Footnote
Supports content creation, image generation, and retrieval augmented generation (RAG) for enhanced model accuracy.
35
What is the main function of Amazon SageMaker?
Comprehensive service for building, training, and deploying custom ML models
## Footnote
Enables custom model development for predictive analytics, large-scale data processing, and real-time inference.
36
What is SageMaker Ground Truth?
A data labeling service to build highly accurate training datasets for machine learning quickly
## Footnote
Features human-in-the-loop labeling and automated data labeling using machine learning models.
37
What feature does SageMaker Canvas provide?
Enables business analysts to build machine learning models and generate accurate predictions without writing code
## Footnote
Offers a no-code interface and automated model generation.
38
What is the purpose of SageMaker Model Monitor?
Monitors deployed models in production for data and model quality issues
## Footnote
Automatically detects and alerts on potential problems.
39
Define inference in machine learning.
The process of generating outputs from new inputs using a trained model.
40
What is the significance of temperature in inference parameters?
Controls the randomness of responses
## Footnote
Higher values increase diversity.
41
What does the term 'Bias Mitigation' refer to?
Address biases present in training data to ensure ethical and fair outcomes
## Footnote
Strategies include analyzing and correcting biases in training data.
42
What is the role of a Vector Database?
Store data as numerical vectors and facilitate efficient lookups
## Footnote
Enhances model capabilities by providing relevant data.
43
What are the core dimensions of Responsible AI?
* Fairness
* Explainability
* Robustness
* Privacy and Security
* Governance
* Transparency
44
What does the Shared Responsibility Model in AWS entail?
AWS ensures the physical security of data centers, hardware, and networking components
## Footnote
AWS manages the underlying infrastructure, including physical servers and operating systems.
45
What is the purpose of evaluation metrics in machine learning?
Assess the performance of models based on specific tasks and requirements
## Footnote
Metrics can include accuracy, precision, recall, F1 score, RMSE, MAP, and MAE.
46
What is Few-Shot Prompting?
Providing a few examples to help the LLM understand and generate responses better.
47
What is a prompt in the context of language models?
A specific set of inputs provided to a language model to guide it in generating an appropriate response or output.
48
What is the function of SageMaker Model Registry?
A centralized repository to store, version, and manage machine learning models
## Footnote
Supports model versioning and integration with deployment pipelines.
49
What is the purpose of Amazon SageMaker Experiments?
A tool to organize, track, compare, and evaluate machine learning experiments
## Footnote
Features experiment tracking and lineage tracking.
50
What does the term 'Customization' refer to in the context of AI models?
Modify or extend the pre-trained model to fit specific needs.
51
What is the function of SageMaker Pipelines?
A service to build, automate, and manage end-to-end machine learning workflows
## Footnote
Supports workflow orchestration and model deployment.
52
What does Amazon SageMaker Inference Recommender do?
Helps select the best compute instance and configuration for inference workloads by running benchmark tests
## Footnote
Provides instance type recommendations and configuration testing.
53
What is equity in AI systems?
Ensures equitable treatment across diverse groups
54
What does explainability in AI refer to?
Provides clear reasons for AI decisions
55
Define robustness in the context of AI.
Ensures tolerance to failures and minimizes errors
56
What is the purpose of privacy in AI?
Protects user data and ensures PII is not exposed
57
What does governance mean in AI systems?
Meets compliance and risk management standards
58
Define transparency in AI.
Clearly communicates model capabilities and risks
59
What is the Shared Responsibility Model in AWS security?
Divides responsibilities between AWS and customers regarding security
60
What are AWS responsibilities in the Shared Responsibility Model?
Security of the Cloud, including Infrastructure Security and Service Management
61
What are customer responsibilities in the Shared Responsibility Model?
Service Configuration and Application Security
62
What is the purpose of Identity and Access Management (IAM) in AWS?
Manages access to AWS resources, including user creation, permissions, and MFA
63
Who is the root user in AWS IAM?
The initial account with unrestricted access
64
What are IAM policies?
JSON documents that define permissions for AWS services and resources
65
What is the Principle of Least Privilege in IAM?
Grant only the permissions necessary for completing specific tasks
66
What are IAM groups?
Collections of IAM users with shared permissions
67
What is the benefit of using IAM roles?
Provides temporary access to AWS resources, reducing risk from long-lived credentials
68
Define identity federation in AWS.
Users authenticate through external identity providers to receive temporary AWS credentials
69
What is the purpose of AWS CloudTrail?
Captures API calls and events within AWS
70
What does S3 Block Public Access do?
Blocks public access to S3 buckets and objects
71
What is the purpose of data encryption?
Protects data confidentiality by making it unreadable without decryption keys
72
What is Client-Side Encryption?
Customers encrypt data before sending it to AWS services
73
What does AWS Key Management Service (KMS) do?
Creates, manages, and controls encryption keys
74
What is the purpose of Amazon Macie?
Monitors S3 buckets for data security and identifies sensitive data
75
What is the role of a Virtual Private Cloud (VPC) in AWS?
Creates private networks within AWS
76
What are AI model vulnerabilities?
Risks such as training data access, adversarial inputs, and model inversion
77
What is the purpose of Amazon SageMaker Model Monitor?
Monitors model performance and data quality in production
78
What does AWS Audit Manager do?
Maps compliance requirements to AWS usage data and generates assessment reports
79
What is AWS Config used for?
Monitors resource configurations and compliance
80
What is the purpose of Amazon Inspector?
Assesses security vulnerabilities in applications and containers
81
What does AWS Trusted Advisor evaluate?
AWS environment against best practices for cost optimization, performance, security, and operational excellence
82
What is the function of AWS Glue DataBrew?
Provides visual data preparation and data quality management tools
AWS AI Certified Practitioner Study Notes
flashcards
Decks in class (6)
# Cards
