A* algorithm - A* is a search algorithm that finds the lowest-cost path from a starting point to a goal by combining two numbers for each option: the known cost so far and...
A/B testing - A/B testing is a randomized experiment that compares two versions of a product, process, or experience by assigning eligible units to different conditions...
A-D conversion - A-D conversion means analog-to-digital conversion: turning a continuously varying physical signal into a sequence of digital values.
A-law algorithm - A-law is a companding law used to represent audio samples with nonuniform quantization, notably in the G.711 telephony standard.
A-optimal design - A-optimal design is an experimental-design criterion that chooses observations to minimize the sum, or equivalently the average for a fixed number of...
AAC - AAC stands for augmentative and alternative communication: ways of communicating that supplement or replace speech when needed.
Abductive reasoning - Abductive reasoning generates or selects plausible explanations for observations.
Abess - Abess, short for Adaptive Best Subset Selection, is a method and software library for selecting a small subset of predictors while fitting a statistical model.
Ablation - An ablation is a controlled experiment that removes one part of a model or system and measures what changes, to learn how much that part actually contributes.
Absolute error - Absolute error is the size of the difference between a predicted or estimated numerical value and its reference value, without regard to direction.
Absolute error loss - Absolute error loss measures a numerical prediction's error as the absolute difference between its predicted and reference values: L(y,y-hat)=|y-y-hat|.
Absolute frequency - Absolute frequency is the number of times a specified value, category, or event occurs in a defined dataset.
Absolute position - Absolute position identifies where a token occurs within a sequence using its own position index, such as first, second, or third.
Abstract word - An abstract word expresses a meaning that is less directly tied to a specific sensory or physical experience, such as an idea, relation, quality, or mental...
Abstraction - Abstraction is the deliberate representation of something using selected features while leaving out other details.
Abstraction hierarchy - An abstraction hierarchy is an arrangement of representations at different levels of detail, connected so that a problem can be considered broadly and then...
Abstractive Summarization - Abstractive summarization is generating a summary in new words that capture the meaning of the source, rather than selecting and copying the source's own...
ABX test - An ABX test is a discrimination test in which a participant compares known references A and B with an unidentified sample X and decides whether X matches A...
Accelerated failure time model - An accelerated failure time model, or AFT model, relates explanatory variables to the time until an event by stretching or compressing the time scale.
Accented syllables - Accented syllables are syllables that bear a pitch accent in an utterance and are perceived as prosodically prominent.
Acceptable quality limit - Acceptable quality limit is a term commonly associated with AQL, a quality level used to select or describe acceptance-sampling schemes.
Acceptance probability - Acceptance probability is the probability that an algorithm accepts a proposed candidate, given its current state and decision rule.
Acceptance sampling - Acceptance sampling is a procedure for deciding what to do with a lot or batch by inspecting a sample rather than every item.
Access Control - Access control is the process of permitting or denying access to resources according to defined rules.
Access control list - An access control list, or ACL, is a set of entries describing access permissions associated with a resource.
Access Manager - In database architecture, an access manager is the part of a database management system that turns an execution plan into operations that access stored data.
Access Method - In database systems, an access method is a set of software routines that provides operations for a particular table or index structure.
Accessibility relations - Accessibility relations are connections between possible worlds or states in a formal model, specifying which alternatives are relevant from a given world.
Accountability - Accountability in AI means that identifiable people and organizations can be called upon to explain and justify decisions within their roles, demonstrate...
Accumulated local effects - Accumulated local effects, or ALE, describe how a fitted model's predictions change with a feature by averaging local prediction differences and...
Accumulating eligibility trace - An accumulating eligibility trace is a temporary learning variable that grows when a state, state-action pair or feature is activated and decays between...
Accuracy - Accuracy is the proportion of evaluated predictions that match the correct labels under a stated scoring rule.
Accuracy barrier - An accuracy barrier is a proposed limit to predictive accuracy arising from variability or uncertainty in what is being predicted.
ACE model - The ACE model is a statistical model that partitions variation in an observed trait into additive genetic variance (A), shared environmental variance (C)...
ACID transactions - ACID transactions are database transactions designed around four properties: atomicity, consistency, isolation and durability.
Acoustic model - Definition: An acoustic model is a component of a speech-recognition system that relates recorded speech signals to linguistic units or their internal states.
Acquiescence bias - Acquiescence bias is a response tendency to agree with statements regardless of their specific content.
Action - In AI, an action is anything an agent can do that changes its environment or its relationship to it: moving, sending, writing, booking, buying.
Action cost function - An action cost function assigns a numerical cost to taking an action in a planning or search problem.
Action model learning - Action model learning is the process of learning descriptions of what actions require and how they change an environment.
Action monitoring - Action monitoring is the practice of checking whether the conditions required for the next planned action still hold before attempting it, and detecting...
Action schema - An action schema is a parameterized description of a type of action in an AI planning model.
Action space - An action space is the set of actions an agent can choose from in a decision-making environment.
Activation - In a neural network, an activation is a numerical value produced by a unit or layer for a particular input, commonly after applying an activation function...
Activation function - An activation function transforms a neural network's intermediate values before they pass onward or become outputs.
Activation Layer - An activation layer is a neural-network component that applies an activation function to incoming values.
Active learning - Active learning is a machine-learning approach where the system chooses which examples it wants labeled next, instead of learning from a fixed or random set.
Active learning agent - An active learning agent is a system that improves itself by choosing which examples it wants to be taught, instead of passively accepting whatever data it...