AWS AI Practitioner AIF-C01 Domain 1 Review

A comprehensive set of vocabulary flashcards covering the fundamentals of AI and ML for the AWS AIF-C01 exam, including terminology, learning methods, and the ML lifecycle.

Artificial Intelligence (AI)

Machines mimicking Human thinking and intelligence; the broad category containing ML, DL, and GenAI.

Machine Learning (ML)

Systems that learn from data without being explicitly programmed; a subset of AI.

Deep Learning (DL)

A subset of ML that uses neural networks with multiple layers (brain-like nodes) to perform complex tasks.

Generative AI (GenAI)

A subset of Deep Learning that creates new content such as text, images, or audio.

Neural Networks

Brain-inspired computing systems comprised of connected nodes called "neurons".

Computer Vision

The field of AI that enables machines to see and understand images and videos.

Natural Language Processing (NLP)

The field of AI that enables machines to understand human language, including text and speech.

Large Language Models (LLMs)

AI models trained on massive text data, such as ChatGPT or Amazon Titan.

Model

A trained algorithm that makes predictions; sometimes referred to as the "brain" after training.

Algorithm

A step-by-step recipe used by computers for solving a problem.

Training

The process of teaching a model using specific data.

Inferencing

The phase where a model makes predictions on new, unseen data based on its training.

Bias

Unfair preferences in AI decisions leading to discriminatory results, often caused by bad input data.

Fairness

The principle of ensuring equal treatment across all groups and no discrimination in AI output.

Underfitting

A model fit issue where the model fails to learn the patterns in the training data sufficiently.

Overfitting

A model fit issue where the model memorizes the training answers too closely and cannot generalize to new data.

Batch Inferencing

Processing many predictions at once (in bulk); used for non-time-sensitive tasks and large volumes of data.

Real-time Inferencing

Generating instant predictions on demand with low latency; used for time-sensitive, user-facing applications.

Labeled Data

Data that includes known answers or tags; used primarily for supervised learning.

Unlabeled Data

Raw data without any tags or answers; used primarily for unsupervised learning to find patterns.

Structured Data

Organized data in a fixed format, such as SQL databases, spreadsheets, or CSV files.

Unstructured Data

Data with no predefined format, such as emails, images, videos, and audio; accounts for approximately $80\%$ of the world's data.

Supervised Learning

A learning method where the model learns from labeled examples with a known mapping between input $(X)$ and output $(Y)$.

Classification

A supervised learning task that predicts a specific category or label (e.g., Spam vs. Not Spam).

Regression

A supervised learning task that predicts a continuous numerical value (e.g., house prices or temperature).

Unsupervised Learning

A learning method where the model finds hidden patterns or structures in unlabeled data without a "teacher" or known answers.

Clustering

An unsupervised learning method that groups similar items together based on common characteristics.

Association

An unsupervised learning method that finds rules between items, such as "people who buy X also buy Y".

Reinforcement Learning (RL)

A learning method based on trial and error where an agent performs actions in an environment to maximize rewards.

Data Preparation

The most time-consuming step of the ML lifecycle ($60-80\%$ of total time) involving cleaning, transforming, and splitting data.

Model Evaluation

The step of testing a model with unseen data to check metrics like Accuracy, Precision, Recall, and F1 Score.

Model Drift

The phenomenon where a model's accuracy or performance drops over time, requiring retraining.

Data Drift

The phenomenon where the underlying data changes over time, affecting the model's relevance.