What is the third phase of building a machine learning model? Data collection Training the model Selecting an appropriate algorithm Model evaluation

Selecting an appropriate algorithm

What is the fourth phase of building a machine learning model? Training the model on prepared data Data preparation Model evaluation Data collection

Training the model on prepared data

AWS - Certified AI Practitioner

Flashcard

Learn Mode

Match

Library

Create

Flashcards

Library

Match (Coming Soon)

Computer Science - Software Engineering

user_kambeorv Created by 7 mon ago

Cards in this deck(100)

What is the broad field that encompasses the development of intelligent systems capable of performing tasks that typically require human intelligence?

What type of AI focuses on understanding and building methods that enable machines to learn and improve performance on tasks?

Which AI technique uses the concept of neurons and synapses similar to how the human brain is wired?

What is the subset of deep learning that can adapt models without retraining and is capable of generating new data?

What is the first phase of building a machine learning model?

What is the second phase of building a machine learning model?

What is the third phase of building a machine learning model?

What is the fourth phase of building a machine learning model?

What is the fifth phase of building a machine learning model?

What type of dataset includes instances accompanied by labels or target variables representing the desired output?

What type of dataset consists only of input features without any corresponding output or classification?

What type of data is organized in a predefined manner, typically in tables or databases, suitable for traditional ML algorithms?

What type of data lacks a predefined structure or format, requiring advanced ML techniques to extract meaningful patterns?

Which learning type involves training ML algorithms on labeled data to learn a mapping function for predicting outputs?

Which learning type involves training ML algorithms on unlabeled data to discover inherent patterns or structures?

Which learning type involves providing feedback in the form of rewards or penalties to improve decision-making over time?

What is the process of using a model's learned information to make predictions or decisions, which can be real-time or batch?

What type of inferencing involves analyzing a large amount of data at once to provide a set of results, often used in data analytics?

What type of inferencing requires quick decision-making in response to new information, crucial for applications like chatbots or self-driving cars?

What are networks made up of nodes organized into layers, capable of identifying patterns by adjusting connections between nodes?

What field of AI enables computers to interpret and understand digital images and videos, often used with deep learning?

What branch of AI deals with the interaction between computers and human languages, used for tasks like text classification and sentiment analysis?

What are models pre-trained on internet-scale data, used in generative AI for tasks like text generation and image generation?

What lifecycle includes data selection, pre-training, optimization, evaluation, deployment, and feedback for continuous improvement?

What are powerful models trained on vast amounts of text data, capable of understanding and generating human-like text?

What are the basic units of text that LLMs process, which can be words, phrases, or individual characters?

What are numerical representations of tokens that capture their meaning and relationships with other tokens in a model?

What DL architecture starts with pure noise and gradually adds meaningful information to generate a coherent output?

What process involves gradually introducing noise to an input image until only the noise is left?

What process involves gradually denoising a noisy image until a new image is generated?

What models can process and generate multiple modes of data simultaneously, learning connections between modalities like images and text?

What type of generative model involves two neural networks competing against each other in a zero-sum game framework?

What type of generative model combines ideas from autoencoders and variational inference to generate new data?

What is the fastest and lowest cost FM optimization technique that involves developing and optimizing prompts?

What FM optimization technique involves a supervised learning process with specific, smaller datasets to modify model weights?

What FM optimization technique uses examples of how the model should respond to specific instructions?

What technique provides human feedback data to align a model better with human preferences?

What FM optimization technique supplies domain-relevant data as context to produce responses based on that data?

What supervised learning use case is used to assign labels or categories to new, unseen data instances?

What supervised learning use case is used for predicting continuous or numerical values based on input variables?

What unsupervised learning use case is used to group data based on similar features or distances between data points?

What unsupervised learning use case is used to reduce the number of features or dimensions in a dataset?

What refers to practices ensuring AI systems are transparent and trustworthy while mitigating potential risks?

What can cause ML models to make prejudiced predictions, leading to inadvertent harm and negative stereotypes?

What is the possibility of generating content that is offensive, disturbing, or otherwise inappropriate in generative AI?

What are assertions or claims that sound plausible but are verifiably incorrect, particularly in LLMs?

What ensures that AI systems do not lead to discriminatory outputs?

What is the ability of an AI model to clearly explain or justify its internal mechanisms and decisions?

What involves building and operating AI systems to protect data from theft and exposure?

What is the practice of communicating information about an AI system to help stakeholders make informed choices?

What ensures an AI system operates reliably, even with unexpected situations, uncertainty, and errors?

What set of processes is used to define, implement, and enforce responsible AI practices within an organization?

What involves developing algorithms, models, and systems to be responsible, safe, and beneficial for society?

What framework monitors and guides an AI system's behavior to align with human values and intent?

What approach creates products and services that are intuitive, easy to use, and meet user needs?

What lifecycle includes business goal identification, ML problem framing, data processing, and model development?

What dataset is used to ensure that a model generalizes to data it has not seen?

What dataset is used for final tests to ensure that a model's predictive quality meets standards?

What occurs when a model performs well on training data but poorly on evaluation data due to memorization?

What occurs when a model performs poorly on training data due to inability to capture input-output relationships?

What tool helps classify why and how a model gets something wrong by examining TP, TN, FP, and FN answers?

What metric is calculated as (TP + TN) / (TP + TN + FP + FN) to measure model performance?

What metric removes negative predictions, focusing only on positive predictions, calculated as TP / (TP + FP)?

What metric looks at the proportion of correct sets identified as positive, calculated as TP / (TP + FN)?

What probability curve represents the degree of separability, showing TP vs FP at various thresholds?

What metric determines the prediction from the model and compares the difference between prediction and actual outcome?

What metric explains the fraction of variance accounted for by the model?

What involves deploying and hosting ML models on your own infrastructure, either on premises or in the cloud?

What are cloud-based services providing a fully managed environment for deploying and hosting ML models as APIs?

What combines people, technology, and processes to deliver collaborative ML solutions, operationalizing the ML lifecycle?

What involves tracking changes to assets like data, code, and models, ensuring systems can be rolled back to previous versions?

What involves automating various stages in the ML pipeline for repeatability, consistency, and scalability?

What extends validation and testing of code to data and models, deploying newly trained models and monitoring metrics?

What structured process reviews, validates, and approves models before deployment, checking for fairness, bias, and ethics?

What are the capabilities of generative AI, including adaptability, responsiveness, simplicity, and creativity?

What are the challenges of generative AI, including regulatory violations, social risks, and data security concerns?

What lifecycle involves defining a use case, selecting a foundation model, improving performance, and deploying the application?

What is the gold standard of FM evaluation, though it can be time-consuming and expensive?

What are curated collections of data designed specifically for evaluating the performance of LLMs or other AI systems?

What is a quick and scalable way to evaluate FM performance?

What set of metrics is used for evaluating automatic summarization and machine translation systems?

What metric evaluates the quality of machine-generated text, especially in the context of machine translation?

What metric evaluates the semantic similarity between a generated text and one or more reference texts?

What parameters can be used to limit or influence the model response when interacting with foundation models?

What parameters influence the variation in generated responses by limiting outputs to more likely outcomes?

What parameter controls the randomness or creativity of the model's output, set between 0 and 1?

What parameter controls the diversity of text by limiting the number of words the model can choose from?

What parameter limits the number of words to the top most probable words, regardless of their percent probabilities?

What parameters control the maximum length of generated output and specify stop sequences for generation?

What determines the maximum number of tokens the model can generate during inference to prevent excessive output?

What are specific tokens or sequences that signal the model to stop generating further output?

What technique involves presenting a task to a generative model without providing examples or explicit training?

What technique involves providing a language model with contextual examples to guide its understanding for a task?

What technique divides intricate reasoning tasks into smaller steps, using zero-shot or few-shot prompting?

What refers to the intentional introduction of malicious or biased data into a model's training dataset?

What technique influences the outputs of generative models by embedding specific instructions within prompts?

What refers to the risk of exposing sensitive information to a generative model during training or inference?

What refers to the unintentional disclosure or leakage of prompts or inputs used within a model?

What refers to the practice of modifying or circumventing constraints in a generative model to gain unauthorized access?

What key functions include intermediary operations, actions launch, and feedback integration in AI systems?

Ask Our AI Tutor

Get Instant Help with Your Questions

Need help understanding a concept or solving a problem? Type your question below, and our AI tutor will provide a personalized answer in real-time!

How it works

Ask any academic question, and our AI tutor will respond instantly with explanations, solutions, or examples.

Get Started

Browse questions and discover topic-based flashcards
Practice with engaging flashcards designed for each subject
Strengthen memory with concise, effective learning tools

Discover By Topic

AWS - Certified AI Practitioner

Related Decks