Brainscape's Knowledge GenomeTM


Top Flashcards (Certified)
All Flashcards

AI Practitioner > AI Practitioner > Flashcards

AI Practitioner Flashcards

1
Q

Model Artifacts

A

Artifacts produced during model training, consisting of trained parameters, model definition, and metadata, often stored in Amazon S3.

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

Inference Code

A

Software that implements the model by reading the model artifacts and making it deployable for inference tasks.

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

Real-Time Inference

A

An inference type where an endpoint is always available to accept requests, suitable for low-latency, high-throughput tasks.

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

Batch Inference

A

An inference type suitable for offline processing, where large amounts of data are processed upfront and a persistent endpoint is not needed.

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

Supervised Learning

A

A machine learning style where models are trained on pre-labeled data, with both input and desired output specified.

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

Unsupervised Learning

A

A machine learning style that works with unlabeled data, focusing on recognizing patterns and grouping data into clusters.

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

Reinforcement Learning

A

A machine learning method focused on autonomous decision-making by an agent, which learns through trial and error by receiving rewards for goal-oriented actions.

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

Amazon SageMaker Ground Truth

A

A service from Amazon that helps label training data for supervised learning, often leveraging Amazon Mechanical Turk for crowdsourcing.

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

Amazon Mechanical Turk

A

A crowdsourcing platform that provides access to a global pool of affordable labor, often used in labeling data for machine learning models.

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

Clustering

A

A process in unsupervised learning where data is grouped based on patterns, useful in anomaly detection and pattern recognition.

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

Anomaly Detection

A

A use case for unsupervised learning where irregularities in data, such as outliers, are identified for further analysis.

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

Reinforcement Learning Agent

A

The entity in reinforcement learning that takes actions within an environment to achieve specific goals, learning through trial and error.

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

AWS DeepRacer

A

A reinforcement learning platform where users teach a model race car (the agent) to navigate a track (the environment) by taking actions to stay on course.

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

Exploratory Approach (Reinforcement Learning)

A

A learning approach where the agent explores actions without knowing the outcome, with successful actions being reinforced for goal achievement.

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

End Goal (Reinforcement Learning)

A

A predetermined objective in reinforcement learning that the agent works towards by refining its actions through trial and error.

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

Artificial Intelligence (AI)

A

The field of computer science dedicated to solving cognitive problems like learning, creation, and image recognition, aiming to create self-learning systems that derive meaning from data.

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

Machine Learning (ML)

A

A branch of AI that focuses on using data and algorithms to imitate the way humans learn, gradually improving accuracy to make predictions.

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

Deep Learning

A

A subset of machine learning inspired by the human brain, using layers of neural networks to recognize speech, images, and more.

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

Inference

A

A prediction made by an AI model, essentially an educated guess with a probabilistic result.

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

Regression Analysis

A

A technique used in AI to process historical time series data and predict future values.

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

Natural Language Processing (NLP)

A

A branch of AI that allows machines to understand, interpret, and generate human language in a natural way.

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

Generative AI

A

AI technology capable of generating original content such as text, images, videos, and music, based on a given prompt.

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

Computer Vision

A

AI technology used to process images and video for tasks like object identification, classification, and facial recognition.

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

Anomaly Detection

A

The process of recognizing deviations from expected patterns in data, often used in fraud detection or identifying system failures.

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
25
Amazon Bedrock
A platform that supports generative AI tasks such as generating content based on prompts, like creating song lyrics or stories.
26
Chatbot
An AI application that uses NLP to engage in conversations with users, often used in customer service or for booking systems.
27
Time Series Data
Historical data points collected over time, which AI models use to identify patterns and make future predictions.
28
Alexa
An AI-powered voice assistant that uses NLP to respond to user questions and perform tasks.
29
Amazon SageMaker
A platform that allows developers to build, train, and deploy machine learning models at scale.
30
Fraud Detection
An AI application used by financial institutions to detect anomalous activity and prevent fraudulent transactions.
31
Computer Vision Example
A model that can detect scratches on surfaces or missing components on circuit boards using image processing.
32
HR AI Applications
AI used to process resumes and match candidates to job roles, improving hiring efficiency.
33
Product Recommendation
An AI-driven system that uses shopping history to suggest products to customers.
34
Discovery (AI Example)
A media platform using AI to recommend personalized content based on viewing history.
35
Customer Support Translation
An AI system that translates between languages in real-time during customer support interactions, such as from English to Spanish.
36
Taxi Demand Forecasting
An AI application that helps taxi companies position cars based on forecasted customer demand.
37
Pandemic Prediction
AI systems used by agencies like the CDC to predict outbreaks and pandemics, aiding in the distribution of resources.
38
AI in Manufacturing
The use of AI with computer vision to monitor assembly lines, maintain product quality, and predict equipment maintenance needs.
39
AI in Medical Diagnosis
AI used to read X-rays and scans to help doctors make faster and more accurate diagnoses.
40
Call Center Monitoring
An AI system that detects deviations in call volumes to identify issues, such as system outages.
41
Machine Learning (ML)
The science of developing algorithms and statistical models that enable computers to perform tasks without explicit instructions by identifying patterns in large datasets.
42
Inference
The process where a trained machine learning model uses new data to make predictions or generate output it hasn't seen during training.
43
Structured Data
Data stored in a table format, such as text files (CSV) or relational databases (RDS, Redshift), with clearly defined rows and columns.
44
Semi-Structured Data
Data that doesn’t fully follow the tabular structure, like JSON files, where features are stored as key-value pairs.
45
Unstructured Data
Data that does not follow any structured model, such as images, videos, or social media posts, typically stored as objects in systems like Amazon S3.
46
Time Series Data
Data that is labeled with timestamps and stored sequentially, often used for models that predict trends over time.
47
Amazon S3 (Simple Storage Service)
Amazon's storage service, which can store any type of data, offers lower cost, virtually unlimited storage capacity, and is a primary source for training data.
48
Features
Input variables (columns in a table, pixels in an image) used to train machine learning models and correlate with expected outputs.
49
Model Parameters
Internal variables in a machine learning model, adjusted during training to improve the model’s ability to produce accurate predictions.
50
Linear Regression
A simple machine learning algorithm that models the relationship between inputs and outputs using a linear equation.
51
Tokenization
A text processing technique that breaks down text into individual units such as words or phrases, commonly used for unstructured data.
52
Training Data
Known data used to train a machine learning model, consisting of features (inputs) and possibly expected outputs.
53
Amazon RDS (Relational Database Service)
Amazon’s service for structured data storage in relational databases, often used as a source for machine learning training data.
54
Amazon Redshift
A fully managed data warehouse service from Amazon, used for analyzing large datasets stored in structured formats.
55
Amazon DynamoDB
Amazon's NoSQL database designed to handle semi-structured data, such as JSON files.
56
Amazon DocumentDB
Amazon's document database with MongoDB compatibility, used for semi-structured data in key-value pairs.
57
Sampling Rate (Time Series Data)
The frequency at which data points are captured in time series data, which can impact data size and model training.
58
Linear Equation
An equation in the form y=mx+b, or for machine learning, an equation used to model the relationship between independent and dependent variables (e.g., h=mw+b).
59
Model Training
The process of adjusting model parameters to reduce errors and improve the accuracy of predictions using known data.
60
Errors (in ML)
The differences between predicted outputs and actual data points, used to refine and improve machine learning models.
61
Object Storage System
A storage system like Amazon S3 that holds unstructured data in the form of objects rather than tables.
62
Machine Learning (ML)
The science of developing algorithms that allow computers to learn from data and make predictions without being explicitly programmed.
63
Supervised Learning
A type of ML where the model is trained on labeled data to make predictions or classifications.
64
Unsupervised Learning
A type of ML where the model is trained on unlabeled data to find hidden patterns or groupings.
65
Reinforcement Learning
An ML method where an agent learns by interacting with its environment and receiving feedback in the form of rewards or penalties.
66
Structured Data
Data that is organized in a tabular format with rows and columns, such as databases or CSV files.
67
Semi-Structured Data
Data that does not have a strict tabular format but includes tags or markers to separate elements, such as JSON or XML.
68
Unstructured Data
Data that does not have a predefined structure, such as images, videos, or social media posts.
69
Amazon S3
Amazon's cloud storage service that stores any type of data with virtually unlimited storage capacity.
70
Inference
The process of making predictions or decisions using a trained ML model on new data.
71
Regression
An ML task that predicts a numerical value based on input features.
72
Classification
An ML task that assigns labels or categories to data points based on input features.
73
Natural Language Processing (NLP)
A field of AI focused on the interaction between computers and human language, enabling tasks like translation and sentiment analysis.
74
Bias
A systematic error in an ML model that occurs when certain data is over or underrepresented.
75
Explainability
The ability to understand and interpret how an AI model makes its decisions.
76
Computer Vision
A field of AI that enables computers to interpret and process visual data like images and videos.
77
Generative AI
A type of AI that can create new content such as text, images, or music, based on learned data.
78
Amazon SageMaker
A fully managed service that provides tools to build, train, and deploy machine learning models at scale.
79
Time Series Data
Data that is indexed over time, often used for trend prediction in ML models.
80
Deep Learning
A subset of ML that uses neural networks with multiple layers (deep neural networks) to process data and find patterns.
81
Model Deployment
The process of putting an ML model into production so it can make real-world predictions or inferences.
82
Amazon Lex
AWS service for building conversational interfaces like chatbots.
83
Amazon Rekognition
AWS service for analyzing images and videos for tasks like object detection and facial recognition.
84
Data Augmentation
A technique used to increase the amount of training data by creating modified versions of the original data.
85
Hyperparameter Tuning
The process of optimizing the parameters that control how an ML algorithm learns.
86
Amazon Lex
Chatbots, like Alexa. Conversational interfaces.
87
Transcribe
Convert speech into text. Transcribe audio recording. Use a custom vocabulary for domain-specific terms.
88
Comprehend
Natural Language Processing. Can identify the language used in text, extract key phrases, and analyze positive/negative sentiment.
89
Textract
Extract data from scanned documents.
90
Rekognition
Computer vision. Analyze images and videos.
91
Kendra
Intelligent search service provider.
92
Personalize
Realtime customer recommendations.
93
DeepRacer
Race car hands on experience with reinforcement learning.
94
CloudWatch
Gather and view metrics that relate to account resources. View the number of API calls to Amazon Bedrock.
95
CloudTrail
Monitor and log API calls in AWS accounts.
96
Amazon Augmented AI (A2I)
Monitoring and human reviews of model quality: Allow for human review of MLs. Automatically builds workflows to allow for human review without burden of building a whole system.
97
AWS AI Service Cards
Part of responsible AI documentation. Provide transparency. Provide a single place to find information on intended use cases, design choices, deployment, optimization best practices. Basic concepts to help customers understand services and features, intended use cases and limitations, responsible AI design considerations, guidance on deployment/optimization.
98
Amazon Q
AI assistant designed for business data. Generative AI virtual assistant that can answer questions, summarize content, generate content, and complete tasks based on the provided data. Does not provide access to FMs and is not open source.
99
Amazon Q Developer
Formerly CodeWhisperer.
100
SageMaker
Amazon's machine learning creation service. Build, train, and deploy custom models. Bring your own machine learning algorithms. Managed API - model hosting services.
101
SageMaker JumpStart
Pre-built foundation models for the most common use cases. Pre-trained models, open source models, FMs for summarization and audit use cases.
102
SageMaker Clarify
FM evaluation: Use metrics such as accuracy, robustness, and toxicity to support your responsible AI initiative. Explainability: Provide insight into your overall model prediction process. Bias detection analysis: by selecting criteria like gender or age. Model Prediction Explanation: Integrated with SageMaker Experiments to provide insight into your overall model prediction process.
103
SageMaker Model Monitor
Monitoring and human reviews: Monitor quality of models in production. Continuous/batch monitoring, on-schedule monitoring. Set alerts: Notify you when there are deviations in model quality.
104
SageMaker Experiments
Provide insight into your overall model prediction process. Help you determine what the impact of incremental changes on model accuracy is.
105
SageMaker Automatic Model Tuning
Hyperparameter tuning: Run many jobs with different hyperparameters and measure each of them.
106
SageMaker Autopilot
Explainability: Provide insights into how ML models make predictions. Determines what contribution certain features or inputs have on the model's output to understand why a model made a prediction after training or use it to provide per-instance explanation during inference.
107
SageMaker Data Wrangler
Rebalance imbalanced data to avoid bias using: Random Undersampling, Random Oversampling, Synthetic Minority Oversampling Technique (SMOTE).
108
SageMaker Feature Store
Create, share, and manage ML development.
109
SageMaker Canvas
Low-code/No-code option to generate predictions without needing to write any code.
110
SageMaker Model Cards
Governance: create records and document details about ML models in a single place. Intended use, risk ratings, training details. Transparency and Explainability: Provides comprehensive, immutable documentation of essential model information.
111
SageMaker Role Manager
Governance: Admins can define user permissions. Control access and security levels for ML activities.
112
SageMaker Model Dashboard
Governance: Keep your team informed on model behavior.
113
SageMaker Ground Truth
Human-in-the-loop feedback. Incorporates human feedback across the ML lifecycle by ranking or classifying responses based on Reinforcement learning from human feedback (RLHF) for fine-tuning.
114
Bedrock
Makes FMs available through an API. Privately customize FMs and deploy.
115
Guardrails for Amazon Bedrom
Generate safeguards based on your use cases and responsible AI policies. Create multiple guardrails for different use cases. Control interactions between users and FMs. Filter undesirable/harmful content. Redact PII. Monitor and analyze user inputs and FM responses.

AI Practitioner (1 decks)

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