AIF-C01-Terms.md

Category	Term	Definition
Custom Language Models	Amazon SageMaker	A comprehensive service offering tools to build, train, and deploy ML models at scale.
AI Concepts	Artificial Intelligence (AI)	AI is the broad field of creating machines or systems that perform tasks requiring human intelligence.
Supervised Learning	CNNs	Primarily used for image and video recognition tasks in deep learning.
NLP	Deep Learning Models	RNNs, Long Short-Term Memory (LSTMs), Transformers (e.g., BERT, GPT).
Supervised Learning	GANs	Used for generating new data samples similar to the training data.
AI Concepts	Generative AI	A subfield of AI focused on generating new content (text, images, audio) similar to the training data.
ML Pipeline	Model Training	Training the ML model with the prepared data.
Evaluation Metrics	Overfitting	When a model performs well on training data but poorly on unseen data.
Unsupervised Learning	Principal Component Analysis (PCA)	Simplifying complex datasets by reducing the number of variables while retaining the most important information
Machine Learning	Reinforcement Learning	A type of ML where an agent learns to make decisions by interacting with an environment and receiving rewards.
Supervised Learning	RNNs	Well-suited for sequential data, like time series or natural language.
NLP	ROUGE	A set of metrics used to evaluate the quality of text summaries.
Few-shot/Multi-shot Learning	SageMaker JumpStart	Open-source models available in SageMaker JumpStart for use in ML tasks.
AWS Services	SageMaker Model Cards	Documentation capturing details, performance metrics, and lineage of ML models.
Machine Learning	Semi-Supervised Learning	Uses a small amount of labeled data and a large amount of unlabeled data to improve learning efficiency.
Machine Learning	Supervised Learning	A machine learning approach where the model is trained on labeled data to make predictions.
Supervised Learning	Support Vector Machines (SVM)	Finds the hyperplane that best separates data into classes for classification and regression.
NLP	Text Generation	Creates new text based on input prompts (e.g., GPT models).
Supervised Learning	Transformers	Advanced neural network architecture used for NLP tasks, e.g., BERT and GPT.
Machine Learning	Unsupervised Learning	A type of machine learning where the model is trained on data without labeled outcomes.
AWS Services	Amazon SageMaker Ground Truth	A service that helps create high-quality labeled datasets for machine learning.
AWS Services	Amazon SageMaker Model Monitor	Monitors ML models in production for data drift, bias, and performance issues.
Open-Source Foundation Models (FM)	Amazon Transcribe	Identifies the language of multiple audio files in batch using Amazon Transcribe.
Batch Language Identification	Amazon Transcribe	Custom models in Amazon Transcribe for enhanced transcription accuracy in specific domains.
SageMaker Inference Options	Asynchronous	Model inference asynchronously, handling long-running predictions.
Model Performance	Batch	Model inference performed in batch mode.
Evaluation Metrics	BLEU	A metric used to evaluate machine-generated translations.
Unsupervised Learning	Clustering	Groups similar data points together without pre-labeled categories.
AWS Services	Data Collection	Collecting and preparing data for model training.
Evaluation Metrics	F1 Score	Combines precision and recall to provide a single performance measure for a classification model.
SageMaker Inference Options	Few-shot Learning	A learning approach where a model is trained with a small amount of labeled data.
Generative AI	Image Generation	Produces images from text descriptions or other images (e.g., DALL-E).
Unsupervised Learning	k-Means Clustering	Partitions a dataset into 'k' groups based on feature similarity.
Supervised Learning	Linear Regression	Models the relationship between a dependent variable and one or more independent variables.
AI Concepts	Machine Learning (ML)	A subset of AI focused on developing algorithms that enable computers to learn from data and make decisions.
NLP	Machine Translation	Translates text from one language to another (e.g., evaluated using BLEU).
ML Pipeline	Model Deployment	Deploying the trained model and monitoring performance.
Few-shot/Multi-shot Learning	Multi-shot Learning	A learning approach that uses multiple examples but fewer than traditional methods.
AI Concepts	Natural Language Processing (NLP)	A subfield of AI focused on enabling computers to read, understand, and generate human language.
Supervised Learning	Neural Networks	Computational models inspired by the human brain, used in deep learning.
SageMaker Inference Options	Real-time	Model inference in real-time for continuous predictions.
Machine Learning	Transfer Learning	Leverages a pre-trained model on one task and fine-tunes it for a related but different task.
Model Performance	Underfitting	Occurs when a model is too simple to capture underlying data patterns.
Unsupervised Learning	Anomaly Detection	Identifies rare items or events that differ significantly from the majority of data.
Generative AI	Audio Generation	Synthesizes new audio, like music or speech.
NLP	BLEU	A metric used to evaluate the quality of machine-generated translations.
Generative AI	BLEU	Evaluates machine-generated text quality against reference text.
Deep Learning	CNNs	Primarily used for image and video recognition tasks in deep learning.
AI Concepts	Computer Vision	AI discipline enabling machines to interpret and understand visual information from the world.
Supervised Learning	Decision Trees	A tree-like model that splits data into branches for predictions.
Reinforcement Learning	Deep Q-Networks	Combines Q-learning with deep neural networks.
Deep Learning	GANs	Generates new data similar to the training data.
NLP	Language Modeling	Predicts the next word or sequence in a sentence (e.g., GPT models).
NLP	Named Entity Recognition (NER)	Identifies and classifies key elements in text.
Reinforcement Learning	Q-Learning	Learns the value of actions in a particular state, without a model of the environment.
Supervised Learning	Random Forest	An ensemble method combining multiple decision trees to improve accuracy.
Deep Learning	RNNs	Well-suited for sequential data, like time series or natural language.
Generative AI	ROUGE	Measures the overlap between machine-generated and reference summaries.
Open-Source Foundation Models (FM)	SageMaker JumpStart	Foundation models provided for rapid development in SageMaker JumpStart.
ML Pipeline	SageMaker Model Cards	Documentation that captures the details, performance metrics, and lineage of machine learning models within SageMaker.
NLP	Sentiment Analysis	Determines the sentiment expressed in text.
NLP	Text Classification	Categorizes text into predefined categories.
NLP	Traditional ML Models	Naive Bayes, Support Vector Machines.
Deep Learning	Transformers	Advanced neural network architecture used for NLP tasks, e.g., BERT and GPT.