AWS AI Flashcards
A company makes forecasts each quarter to decide how to optimize operations to meet expected demand. The company uses ML models to make these forecasts. An AI practitioner is writing a report about the trained ML models. What should they include in the report to meet transparency and explainability requirements for company stakeholders?
A. Code for model training
B. Partial dependence plots (PDPs)
C. Sample data for training
D. Model convergence tables
B
What is the primary purpose of normalization in data preprocessing?
A. To split the data into training and test sets
B. To increase the number of features
C. To ensure all features contribute equally to the model
D. To remove outliers from the data
C
Which Amazon Bedrock API is used for making inference requests?
A. bedrock-agent-runtime
B. bedrock-runtime
C. bedrock
D. bedrock-agent
B
Which method is commonly used to handle imbalanced datasets?
A. Random over-sampling of the minority class
B. Decreasing the number of features
C. Applying PCA
D. Using linear regression
A
What is the role of Amazon Q Apps in Amazon Q Business?
A. To provide hardware configurations
B. To monitor application performance
C. To create and share generative AI-powered apps
D. To develop new AWS services
C
What is the primary objective of reinforcement learning?
A. To reduce dimensionality of data
B. To cluster similar data points
C. To predict future values
D. To maximize cumulative reward
D
How does Amazon Bedrock help in managing multi-step tasks?
A. By using data encryption
B. By using fine-tuning
C. By using agents
D. By using model evaluation
C
Which SageMaker feature helps debug machine learning models?
A. SageMaker Debugger
B. SageMaker Ground Truth
C. SageMaker Data Wrangler
D. SageMaker Autopilot
A
Which machine learning technique is used to handle missing data?
A. Imputation
B. Regression
C. Clustering
D. Classification
A
What is the role of the activation function in a neural network?
A. To normalize the input data
B. To reduce the number of parameters
C. To introduce non-linearity into the model
D. To decrease the learning rate
C
What is the purpose of a confusion matrix in classification?
A. To measure the accuracy of regression models
B. To perform dimensionality reduction
C. To evaluate the performance of a classification model
D. To visualize data distributions
C
What is the purpose of SageMaker Studio Lab?
A. To deploy ML models
B. To provide a free environment for ML experimentation
C. To encrypt data
D. To monitor model drift
B
How can developers test their agents in Amazon Bedrock?
A. Running them on local servers
B. By deploying them in production
C. Integrating with CI/CD pipelines
D. Using the Amazon Bedrock console
D
How can Amazon Bedrock Studio help developers?
A. By providing hardware maintenance
B. By managing financial transactions
C. By monitoring network traffic
D. By offering a rapid prototyping environment
D
How does Amazon Textract assist in document processing?
A. By translating text
B. By detecting sentiment in text
C. By converting text to speech
D. By extracting text and data from documents
D
Which feature helps Amazon Q Business avoid generating incorrect statements?
A. Random response generation
B. Relying on user feedback
C. Manual verification of all responses
D. Confining responses to existing enterprise data
D
What does a ROC curve represent?
A. The error rate of a model
B. The relationship between precision and recall
C. The trade-off between sensitivity and specificity
D. The distribution of a dataset
C
What is a common use of decision trees in machine learning?
A. Data normalization
B. Regression and classification
C. Clustering data points
D. Dimensionality reduction
B
Which evaluation metric is suitable for imbalanced classification problems?
A. Adjusted R-Squared
B. Precision-Recall Curve
C. Mean Squared Error (MSE)
D. Accuracy
B
What is the purpose of a validation set in machine learning?
A. To test the final model performance
B. To visualize the data
C. To train the model
D. To tune hyperparameters
D
Which algorithm is used for clustering in unsupervised learning?
A. Linear regression
B. Decision trees
C. Naive Bayes
D. K-means
D
Which algorithm is best suited for large datasets with many features?
A. Gradient Boosting
B. Naive Bayes
C. Random Forest
D. K-means
C
Which AWS service makes it easy to use foundation models from third-party providers and Amazon?
A. AWS Lambda
B. Amazon Bedrock
C. Amazon EC2
D. AWS SageMaker
B
Which component is essential for extending the capabilities of Amazon Bedrock agents?
A. Static IP addresses
B. Action groups
C. File systems
D. Virtual machines
B
Which API in Amazon Bedrock is used for managing and deploying models?
A. bedrock-agent-runtime
B. bedrock-runtime
C. bedrock-agent
D. bedrock
D
What is the benefit of wrapping instructions in a single tagged section in prompts?
A. It prevents the model from revealing the salted sequence
B. It simplifies model training
C. It enhances data storage
D. It reduces API call costs
A
Which method is used to customize models in Amazon Bedrock for specific tasks?
A. Manual coding
B. Fine-tuning with labeled datasets
C. Unsupervised learning
D. Random training
B
How does Amazon Bedrock ensure data privacy during foundation model tuning?
A. By creating a private copy of the foundation model
B. By disabling data encryption
C. By using public datasets
D. By using internet traffic for data transfer
A
Which feature allows Amazon Q Business to handle complex natural language queries?
A. Predefined response templates
B. Semantic document retriever
C. Regular expressions
D. Basic keyword matching
B
What is the main purpose of Amazon Q?
A. Running SQL queries
B. Generating insights from data
C. Deploying machine learning models
D. Data storage
B
Which AWS service can be used to create scalable endpoints for generative AI models?
A. AWS Glue
B. Amazon SageMaker Hosting
C. Amazon Lex
D. Amazon Comprehend
B
Which component of Amazon SageMaker helps in fine-tuning foundation models for specific tasks?
A. SageMaker Studio
B. SageMaker Autopilot
C. SageMaker JumpStart
D. SageMaker Ground Truth
C
Which AWS service can be integrated with Amazon Bedrock for enhanced search capabilities?
A. Amazon Kendra
B. Amazon S3
C. Amazon Polly
D. Amazon RDS
A
What is overfitting in machine learning?
A. When a model has too few parameters
B. When a model generalizes well to new data
C. When a model underperforms on both training and test data
D. When a model performs well on training data but poorly on test data
D
How does Amazon SageMaker Inference optimize foundation model deployment?
A. By simplifying model architecture
B. By increasing storage capacity
C. By enhancing security features
D. By reducing deployment costs and latency
D
How can developers iterate on their Amazon Bedrock agents?
A. By increasing storage space
B. By adjusting network settings
C. By creating new versions and aliases
D. By deploying directly to production
C
What is the role of Amazon SageMaker Clarify?
A. To deploy machine learning models
B. To provide real-time translations
C. To manage data storage
D. To detect and measure bias in AI models
D
Which AWS service can be used for sentiment analysis of customer feedback?
A. Amazon Polly
B. Amazon Comprehend
C. Amazon Lex
D. Amazon Rekognition
B
How does Amazon Polly enhance user experience?
A. By detecting anomalies in data
B. By providing real-time translation
C. By analyzing video content
D. By converting text into lifelike speech
D
How does AWS PrivateLink enhance the security of data transfers?
A. By providing private connectivity between VPCs and AWS services
B. By disabling encryption
C. By increasing data storage capacity
D. By using public internet
A
What is the benefit of using SageMaker Autopilot with foundation models?
A. Automates data storage
B. Increases token limits
C. Automatically builds, trains, and tunes models
D. Enhances security protocols
C
How does Amazon SageMaker Canvas help business analysts?
A. By encrypting data
B. By tuning hyperparameters
C. By providing a visual interface for ML predictions
D. By enabling model deployment
C
What is the primary function of SageMaker Model Monitor?
A. To automate model training
B. To tune hyperparameters
C. To detect concept drift in deployed models
D. To preprocess data
C
Which AWS service facilitates the orchestration of machine learning workflows for generative AI?
A. Amazon S3
B. Amazon Comprehend
C. AWS Step Functions
D. Amazon Polly
C
What is the role of Amazon Kendra in enhancing generative AI solutions?
A. It provides real-time data streaming
B. It offers intelligent search capabilities
C. It manages data encryption
D. It automates model training
B
What type of tasks can Amazon Bedrock agents automate?
A. Financial reporting
B. Code deployment
C. Customer support interactions
D. Infrastructure management
C
Which AWS service can enhance the interpretability of generative AI models?
A. Amazon SageMaker Clarify
B. Amazon Rekognition
C. Amazon Translate
D. Amazon Lex
A
How does SageMaker Inference optimize the deployment of foundation models?
A. By increasing storage capacity
B. By simplifying model architecture
C. By optimizing the use of accelerators
D. By enhancing security protocols
C
How can you prevent prompt injection attacks in Amazon Bedrock?
A. By using static tokens
B. By increasing token count
C. By implementing dynamic tokenization and validation
D. By ignoring user inputs
C
What is the purpose of cross-validation in machine learning?
A. To cluster data points
B. To visualize data
C. To prevent overfitting
D. To increase the complexity of a model
C
Which capability of Amazon SageMaker helps reduce training time for large models?
A. SageMaker Clarify
B. SageMaker Data Wrangler
C. SageMaker Studio
D. SageMaker HyperPod
D
Which integration allows Amazon Q Business to create and manage Jira issues?
A. Built-in SQL queries
B. Custom plugin
C. Amazon EC2 instances
D. Direct integration with AWS Lambda
B
How can AWS Glue assist in preparing data for generative AI models?
A. By providing real-time translations
B. By providing model deployment services
C. By automating data preparation and ETL processes
D. By managing model monitoring
C
What is a support vector machine (SVM)?
A. A type of neural network
B. An unsupervised learning algorithm
C. A clustering technique
D. A supervised learning algorithm used for classification and regression
D
How does Amazon Bedrock ensure the privacy of data used for customizing foundation models?
A. By encrypting data at rest and in transit
B. By applying strict IAM policies
C. By creating private, customized copies of the base model
D. By using dedicated hardware
C
