1. Fundamentals of AWS AI/ML Services Flashcards
(20 cards)
What is the primary difference between AI and ML?
A) AI is a subset of ML
B) ML is a subset of AI
C) They are completely unrelated fields
D) AI and ML are the same thing
Correct Answer: B
Explanation: ML is a subset of AI. The study guide mentions that understanding the similarities and differences between AI, ML, and deep learning is important.
Which of the following is NOT a type of machine learning?
A) Supervised learning
B) Unsupervised learning
C) Reinforcement learning
D) Diagnostic learning
Correct Answer: D
Explanation: The study guide mentions supervised, unsupervised, and reinforcement learning as types of machine learning. Diagnostic learning is not a standard type of ML.
What type of data is most suitable for training a computer vision model?
A) Tabular data
B) Time-series data
C) Image data
D) Text data
Correct Answer: C
Explanation: Image data is most suitable for computer vision models. The study guide mentions different types of data used in AI models, including image data.
Which AWS service is best suited for natural language processing tasks?
A) Amazon SageMaker
B) Amazon Comprehend
C) Amazon Polly
D) Amazon Transcribe
Correct Answer: B
Explanation: Amazon Comprehend is specifically designed for natural language processing tasks. The study guide lists various AWS managed AI/ML services and their capabilities.
What is the primary purpose of exploratory data analysis (EDA) in the ML development lifecycle?
A) To train the model
B) To deploy the model
C) To understand the characteristics of the data
D) To monitor the model in production
Correct Answer: C
Explanation: EDA is used to understand the characteristics of the data before model training. The study guide mentions EDA as a component of an ML pipeline.
Which of the following is NOT a typical stage in an ML pipeline?
A) Data collection
B) Feature engineering
C) Model training
D) Customer acquisition
Correct Answer: D
Explanation: Customer acquisition is not a typical stage in an ML pipeline. The study guide lists the components of an ML pipeline, which do not include customer acquisition
What does AUC stand for in the context of model performance metrics?
A) Average User Cost
B) Area Under the Curve
C) Automated Universal Calculation
D) Augmented Use Case
Correct Answer: B
AUC stands for Area Under the Curve (specifically, the ROC curve). The study guide mentions AUC as one of the model performance metrics
Which type of learning is most appropriate when you have a large dataset of labeled examples?
A) Unsupervised learning
B) Reinforcement learning
C) Supervised learning
D) Semi-supervised learning
Correct Answer: C
Explanation: Supervised learning is most appropriate when you have labeled data. The study guide describes supervised learning as one of the types of machine learning
What is the main advantage of using pre-trained models?
A) They always perform better than custom models
B) They require less computational resources to train
C) They are always more accurate
D) They can be used immediately without any training data
Correct Answer: D
Explanation: Pre-trained models can be used immediately without training data, which is their main advantage. The study guide mentions pre-trained models as a source of ML models.
Which AWS service is best suited for automating the process of identifying the best hyperparameters for a model?
A) Amazon SageMaker Autopilot
B) Amazon Comprehend
C) Amazon Polly
D) Amazon Transcribe
Correct Answer: A
Explanation: Amazon SageMaker Autopilot is designed for automating the process of finding the best hyperparameters. While not explicitly mentioned in the study guide, it falls under the SageMaker suite of tools.
What does MLOps stand for?
A) Machine Learning Operations
B) Multiple Learning Optimizations
C) Model Learning Objectives
D) Managed Learning Outputs
Correct Answer: A
Explanation: MLOps stands for Machine Learning Operations. The study guide mentions MLOps and its fundamental concepts.
Which of the following is NOT a typical business metric for evaluating ML models?
A) Cost per user
B) Development costs
C) Customer feedback
D) F1 score
Correct Answer: D
Explanation: F1 score is a model performance metric, not a business metric. The study guide distinguishes between model performance metrics and business metrics.
What type of learning is most appropriate when you want an agent to learn from its interactions with an environment?
A) Supervised learning
B) Unsupervised learning
C) Reinforcement learning
D) Transfer learning
Correct Answer: C
Explanation: Reinforcement learning is used when an agent learns from interactions with an environment. The study guide mentions reinforcement learning as one of the types of machine learning.
Which AWS service is best suited for converting text to speech?
A) Amazon Comprehend
B) Amazon Translate
C) Amazon Transcribe
D) Amazon Polly
Correct Answer: D
Explanation: Amazon Polly is designed for text-to-speech conversion. The study guide lists various AWS managed AI/ML services and their capabilities.
What is the primary purpose of feature engineering in the ML development lifecycle?
A) To collect more data
B) To create new features or transform existing ones to improve model performance
C) To evaluate the model’s performance
D) To deploy the model to production
Correct Answer: B
Explanation: Feature engineering involves creating new features or transforming existing ones to improve model performance. The study guide mentions feature engineering as a component of an ML pipeline.
Which of the following is an example of unsupervised learning?
A) Spam detection
B) Image classification
C) Clustering customer segments
D) Predicting house prices
Correct Answer: C
Explanation: Clustering is a typical unsupervised learning task. The study guide mentions unsupervised learning as one of the types of machine learning.
What is the main difference between batch inferencing and real-time inferencing?
A) Batch inferencing is always more accurate
B) Real-time inferencing can only be done on small datasets
C) Batch inferencing processes multiple inputs at once, while real-time inferencing processes individual inputs as they arrive
D) Real-time inferencing is always faster than batch inferencing
Correct Answer: C
Explanation: The main difference is in how inputs are processed. The study guide mentions different types of inferencing, including batch and real-time.
Which AWS service is best suited for managing the entire machine learning lifecycle?
A) Amazon Comprehend
B) Amazon SageMaker
C) Amazon Polly
D) Amazon Translate
Correct Answer: B
Explanation: Amazon SageMaker is designed to manage the entire machine learning lifecycle. The study guide mentions SageMaker multiple times in the context of the ML development lifecycle.
What is the primary purpose of model monitoring in production?
A) To train new models
B) To collect more data
C) To detect issues like model drift or data drift
D) To perform feature engineering
Correct Answer: C
Explanation: Model monitoring in production is primarily used to detect issues like model drift or data drift. The study guide mentions model monitoring as part of MLOps
Which of the following is NOT a typical use case for AI/ML?
A) Fraud detection
B) Recommendation systems
C) Manual data entry
D) Speech recognition
Correct Answer: C
Explanation: Manual data entry is not a typical use case for AI/ML. The study guide lists several real-world AI applications, which do not include manual data entry.
