Александр Чесалов – Глоссариум по искусственному интеллекту: 2500 терминов. Том 2 (страница 2)
Таких примеров «удивительных» терминов мы можем привести еще не мало. Но в своей работе мы не стали тратить время на «суровую действительность» и сместили акцент на конструктивный и позитивный настрой. Одним словом, мы провели для Вас большую работу и собрали более 2500 терминов и определений по машинному обучению и искусственному интеллекту на основе своего опыта и данных из огромного числа различных источников.
2500 терминов и определений.
Много это или мало?
Наш опыт подсказывает, что для взаимопонимания двум собеседникам достаточно знать десяток или, максимум, два десятка определений. Но, когда дело касается профессиональной деятельности, то может получиться так, что мало знать, даже, несколько десятков терминов.
В этой книге приведены самые актуальные термины и определения, по-нашему мнению, наиболее часто употребляемые, как в повседневной работе, так и профессиональной деятельности специалистами самых разных профессий, интересующихся темой «искусственного интеллекта».
Мы очень старались сделать для вас нужный и полезный «инструмент» для вашей работы.
В заключение хочется добавить и проинформировать уважаемого читателя о том, что эта книга является абсолютно открытым и свободным к распространению документом. В случае, если Вы используете ее в своей практической работе, просим Вас делать ссылку на нее.
Многие из терминов и определений к ним, в этой книге, встречаются в сети Интернет. Они повторяются десятки или сотни раз на различных информационных ресурсах (в основном на зарубежных). Тем не менее, мы поставили перед собой цель – собрать и систематизировать самые актуальные из них в одном месте из самых разных источников, нужные из них перевести на русский язык и/или адаптировать, а какие-то и написать заново, исходя из собственного опыта.
Учитывая вышесказанное, мы не претендуем на авторство или уникальность представленных терминов и определений, но, несомненно, мы внесли свой собственный вклад в систематизацию и адаптацию многих из них.
Книга написана, прежде всего, для вашего удовольствия.
Мы продолжаем работу по улучшению качества и содержания текста этой книги, в том числе дополняем ее новыми знаниями по предметной области. Будем вам благодарны за любые отзывы, предложения и уточнения. Направляйте их, пожалуйста, на aleksander.chesalov@yandex.ru
Приятного Вам чтения и продуктивной работы!
Ваши, Александр Чесалов, Александр Власкин и Матвей Баканач.
16.08.2022. Издание первое.
09.03.2023. Издание второе. Исправленное и дополненное.
01.01.2024. Издание третье. Исправленное и дополненное.
Artificial Intelligence glossary
«A»
A/B Testing is a statistical way of comparing two (or more) techniques, typically an incumbent against a new rival. A/B testing aims to determine not only which technique performs better but also to understand whether the difference is statistically significant. A/B testing usually considers only two techniques using one measurement, but it can be applied to any finite number of techniques and measures2.
Abductive logic programming (ALP) is a high-level knowledge-representation framework that can be used to solve problems declaratively based on abductive reasoning. It extends normal logic programming by allowing some predicates to be incompletely defined, declared as adducible predicates3.
Abductive reasoning (also abduction) is a form of logical inference which starts with an observation or set of observations then seeks to find the simplest and most likely explanation. This process, unlike deductive reasoning, yields a plausible conclusion but does not positively verify it. abductive inference, or retroduction4.
Abstract data type is a mathematical model for data types, where a data type is defined by its behavior (semantics) from the point of view of a user of the data, specifically in terms of possible values, possible operations on data of this type, and the behavior of these operations5.
Abstraction — the process of removing physical, spatial, or temporal details or attributes in the study of objects or systems in order to more closely attend to other details of interest6.
Accelerating change is a perceived increase in the rate of technological change throughout history, which may suggest faster and more profound change in the future and may or may not be accompanied by equally profound social and cultural change7.
Access to information – the ability to obtain information and use it8.
Access to information constituting a commercial secret – familiarization of certain persons with information constituting a commercial secret, with the consent of its owner or on other legal grounds, provided that this information is kept confidential9.
Accuracy – the fraction of predictions that a classification model got right. In multi-class classification, accuracy is defined as follows:
In binary classification, accuracy has the following definition:
See true positive and true negative. Contrast accuracy with precision and recall10,11.
Action in reinforcement learning, is the mechanism by which the agent transitions between states of the environment. The agent chooses the action by using a policy12.
Action language is a language for specifying state transition systems, and is commonly used to create formal models of the effects of actions on the world. Action languages are commonly used in the artificial intelligence and robotics domains, where they describe how actions affect the states of systems over time, and may be used for automated planning13.
Action model learning is an area of machine learning concerned with creation and modification of software agent’s knowledge about effects and preconditions of the actions that can be executed within its environment. This knowledge is usually represented in logic-based action description language and used as the input for automated planners14.
Action selection is a way of characterizing the most basic problem of intelligent systems: what to do next. In artificial intelligence and computational cognitive science, «the action selection problem» is typically associated with intelligent agents and animats – artificial systems that exhibit complex behaviour in an agent environment15.
Activation function in the context of Artificial Neural Networks, is a function that takes in the weighted sum of all of the inputs from the previous layer and generates an output value to ignite the next layer16.
Active Learning/Active Learning Strategy is a special case of Semi-Supervised Machine Learning in which a learning agent is able to interactively query an oracle (usually, a human annotator) to obtain labels at new data points. A training approach in which the algorithm chooses some of the data it learns from. Active learning is particularly valuable when labeled examples are scarce or expensive to obtain. Instead of blindly seeking a diverse range of labeled examples, an active learning algorithm selectively seeks the particular range of examples it needs for learning17,18,19.
Adam optimization algorithm it is an extension of stochastic gradient descent which has recently gained wide acceptance for deep learning applications in computer vision and natural language processing20.
Adaptive algorithm is an algorithm that changes its behavior at the time it is run, based on a priori defined reward mechanism or criterion21,22.
Adaptive Gradient Algorithm (AdaGrad) is a sophisticated gradient descent algorithm that rescales the gradients of each parameter, effectively giving each parameter an independent learning rate23.
Adaptive neuro fuzzy inference system (ANFIS) (also adaptive network-based fuzzy inference system) is a kind of artificial neural network that is based on Takagi—Sugeno fuzzy inference system. The technique was developed in the early 1990s. Since it integrates both neural networks and fuzzy logic principles, it has potential to capture the benefits of both in a single framework. Its inference system corresponds to a set of fuzzy IF—THEN rules that have learning capability to approximate nonlinear functions. Hence, ANFIS is considered to be a universal estimator. For using the ANFIS in a more efficient and optimal way, one can use the best parameters obtained by genetic algorithm24.
Adaptive system is a system that automatically changes the data of its functioning algorithm and (sometimes) its structure in order to maintain or achieve an optimal state when external conditions change25.
Additive technologies are technologies for the layer-by-layer creation of three-dimensional objects based on their digital models («twins»), which make it possible to manufacture products of complex geometric shapes and profiles26.
Admissible heuristic in computer science, specifically in algorithms related to pathfinding, a heuristic function is said to be admissible if it never overestimates the cost of reaching the goal, i.e., the cost it estimates to reach the goal is not higher than the lowest possible cost from the current point in the path27.
Affective computing (also artificial emotional intelligence or emotion AI) – the study and development of systems and devices that can recognize, interpret, process, and simulate human affects. Affective computing is an interdisciplinary field spanning computer science, psychology, and cognitive science28.