Кафедра штучного інтелекту (ШІ)
Постійний URI для цієї колекції
Перегляд
Перегляд Кафедра штучного інтелекту (ШІ) за назвою
Зараз показано 1 - 20 з 281
Результатів на сторінку
Варіанти сортування
Публікація An introduction to Knowledge Computing(НПП ЧП «Технологический Центр», 2014) Terziyan, V.; Shevchenko, O.; Golovianko, M.Статья посвящена задачам, связанным с самоуправляемыми и эволюционирующими массивами знаний. Мы предполагаем, что граф самоуправляемых знаний представляет из себя гибрид из явных декларативных знаний о себе и явных процедурных знаний. Предложено расширение к традиционной RDF модели, описывающей граф знаний в соответствии со стандартами Semantic Web. Предложена концепция Executable Knowledge и Knowledge Computing, основанная на добавлении исполняемых свойств к традиционно используемым (data type и object type) типам свойств в рамках RDF модели. Статья также представляет пилотную реализацию Executable Knowledge, как модуля для Protege.Публікація Automatic sign language translation system using neural network technologies and 3d animation(2023) Shovkovyi, Y.; Grynyova, O.; Udovenko, S.; Chala , L.Упровадження програмних засобів автоматичного сурдоперекладу в процес соціальної інклюзії людей з вадами слуху є важливим завданням. Соціальна інклюзія для осіб із вадами слуху є нагальною проблемою, яку необхідно вирішувати з огляду на розвиток IT-технологій та законодавчі ініціативи, що забезпечують права людей з інвалідністю та їхні рівні можливості. Сказане обґрунтовує актуальність дослідження асистивних технологій у контексті програмних засобів, таких як процес соціального залучення людей з важкими порушеннями слуху в суспільство. Предметом дослідження є методи автоматизованого сурдоперекладу із застосуванням інтелектуальних технологій. Мета роботи – розроблення та дослідження методів автоматизації сурдоперекладу для поліпшення якості життя людей з вадами слуху відповідно до «Цілей сталого розвитку України» (в частині «Скорочення нерівності»). Основними завданнями дослідження є розроблення й тестування методів перетворення жестової мови в текст, перетворення тексту в жестову мову, а також автоматизації перекладу з однієї жестової мови іншою жестовою мовою із застосуванням сучасних інтелектуальних технологій. Для розв’язання цих завдань використовувались методи нейромережного моделювання та 3D-анімації. Унаслідок дослідження здобуто такі результати: виявлено основні проблеми й завдання соціальної інклюзії для людей з вадами слуху; здійснено порівняльний аналіз сучасних методів і програмних платформ автоматичного сурдоперекладу; запропоновано й досліджено систему, що об’єднує метод SL-to-Text; метод Text-to-SL з використанням 3D-анімації для генерації концептів жестової мови; метод генерації 3D-анімованого жесту з відеозаписів; метод реалізації технології Sign Language1 to Sign Language2. Для розпізнавання жестів застосовано модель згорткової нейронної мережі, що навчається за допомогою імпортованих і згенерованих системою датасетів відеожестів. Навчена модель має високу точність розпізнавання (98,52 %). Створення 3D-моделі для відображення жесту на екран і його оброблення відбувалися у середовищі Unity 3D. Структура проєкту, виконавчих і допоміжних файлів, що застосовуються для побудови 3D-анімації для генерації концептів жестової мови містить: файли обробників подій; результати відображення, що мають інформацію про положення відслідкованих точок тіла; файли, що зберігають характеристики матерій, які були додані до тих чи інших точок відображення тіла. Висновки: запропоновані методи автоматизованого перекладу мають практичну значущість, що підтверджують демоверсії програмних застосунків Sign Language to Text і Text to Sign Language. Перспективним напрямом подальших досліджень з окресленої теми є вдосконалення методів SL1-to-SL2, створення відкритих датасетів відеожестів, залучення науковців і розробників для наповнення словників концептами різних жестових мов.Публікація Autonomous Agent's Behavior Model Based on Automata Theory(ХНУРЕ, 2018) Філатов, В. О.; Семенець, В. В.; Filatov, V. O.; Semenets, V. V.This paper is devoted to a problem of management the corporate system's information space. The management is carried out on the technology of the program agents. The approach to construction the program agent's model on a basis of frame structure is considered. The frame is fixed as universal extension system. For the decision of concrete tasks the functional modules - slots can be added in it. Base structure for model of behavior of the program agents in information environment is the model based on the finite-state machine. The offered models can be used for development the systems for administration of information resources in the allocated computing systems.Публікація Cell tracking improved through machine learning(ХНУРЕ, 2024) Kupriianov, S.The work is devoted to assessing the challenges and advancements in microscopy data analysis, focusing on cell tracking, object detection, lineage reconstruction, and the application of machine learning. It highlights the difficulties in accurate lineage reconstruction, the importance of metrics in assessing tracking algorithms, and the role of machine learning in proposing hypotheses and understanding complex biological systems.Публікація CLIPTraVeLGAN for Semantically Robust Unpaired Image Translation(2023) Bodyanskiy, Y.; Ryabova, N.; Lavrynenko, R.In this paper a novel approach for semantically robust unpaired image translation is presented. CLIPTraVeLGAN replaces the Siamese network in TraVeLGAN with a contrastively pretrained language-image model (CLIP) with frozen weights. This approach significantly simplifies the model selection and training process of TraVeLGAN, making it more robust and easier to use.Публікація Data mining in relational systems(ХНУРЕ, 2020) Філатов, В. О.; Семенець, В. В.; Золотухін, О. В.The subject of the research is methods of relational database mining. The purpose of the research is to develop scientificallygrounded models for supporting intelligent technologies for integrating and managing information resources of distributed computing systems. Explore the features of the operational specification of the relational data model. To develop a method for evaluating a relational data model and a procedure for constructing functional associative rules when solving problems of mining relational databases. In accordance with the set research goal, the presented article considers the following tasks: analysis of existing methods and technologies for data mining. Research of methods for representing intelligent models by means of relational systems. Development of technology for evaluating the relational data model for building functional association rules in the tasks of mining relational databases. Development of design tools and maintenance of applied data mining tasks; development of applied problems of data mining. Results: The analysis of existing methods and technologies for data mining is carried out. The features of the structural specification of a relational database, the formation of association rules for building a decision support system are investigated. Information technology has been developed, a methodology for the design of information and analytical systems, based on the relational data model, for solving practical problems of mining, practical recommendations have been developed for the use of a relational data model for building functional association rules in problems of mining relational databases, conclusion: the main source of knowledge for database operation can be a relational database. In this regard, the study of data properties is an urgent task in the construction of systems of association rules. On the one hand, associative rules are close to logical models, which makes it possible to organize efficient inference procedures on them, and on the other hand, they more clearly reflect knowledge than classical models. They do not have the strict limitations typical of logical calculus, which makes it possible to change the interpretation of product elements. The search for association rules is far from a trivial task, as it might seem at first glance. One of the problems is the algorithmic complexity of finding frequently occurring itemsets, since as the number of items grows, the number of potential itemsets grows exponentially.Публікація Face recognition problems for artificial intelligence in case with social medias(ХНУРЕ, 2020) Логвінова, К. В.Nowadays, a certain kind of experience in the era of information technology and information technology, the problem of creating “smart” technology is especially urgent. On the “smart” technologies on the current day, you can rozumіti cars with programs of any complexity, all kinds of pralny machines, especially the robbery mode before the robots, which can make pictures and cherubs by co-workers.Публікація Formation of alternative approaches to noise generation in gan networks(ХНУРЕ, 2024) Bilokon, V. A.This work discusses the problem of developing alternative approaches to generating noise in generative adversarial networks (GAN). Various noise generation techniques are important for training GAN networks as they help improve the quality of the generated data and the stability of training. This article provides an overview of current noise generation methods and discusses an approach based on the Pandas library in the Python programming language for generating, storing, and mixing noises.Публікація Good practices of Industry 4.0 in Ukraine(2022) Golovianko, M.; Gryshko, S.; Titova, L.; Filatov, V.There are examples and good practices of using AI and developing Industry 5.0 all over Europe. This desk research is made by the team from Kharkiv National University of Radio Electronics within Erasmus+ JoInME Project “JoInt Multidisciplinary training program on Entrepreneurship in the field of artificial intelligence for industry 5.0” to identify the winning practices from Ukrainian startups or companies working in the Industry 4.0/5.0 and using AI.Публікація Kernel principal component analysis in data stream mining tasks(2016) Bodyanskiy, Ye. V.; Deineko, A. O.; Eze, F. M.; Shalamov, M. O.Currently, self-learning systems of computational intelligence [1, 2] and, above all , artificial neural networks (ANN ), that tune their parameters without a teacher on the basis of the self-learning paradigm [3], are widely used in solving various problems of Data Mining, Exploratory Data Analysis etc. Among these tasks, most frequently encountered in the Text Mining, Web Mining, Medical Data Mining, it be can mentioned the problem of compression of large data sets, for whose solution principal component analysis (PCA) is widely used, which consists in the orthogonal projection of input data vectors from the original n-dimensional space in the m- dimensional space of reduced dimensionalityПублікація Llm як інструмент для емоційної нейтралізації тексту(ХНУРЕ, 2024) Білоконь, Б. О.Emotionality in communication can have both positive and negative effects, depending on how it is used and controlled. The aim of this paper is to investigate the effectiveness and accuracy of emotional neutralization of texts using LLM, while preserving semantics and main meaning. The study uses prompt engineering methods to build queries whose answers will have the best results.Публікація Multiagent Approach for Managing the Information Space of Corporate Systems(ХНУРЕ, 2018) Філатов, В. О.; Семенець, В. В.; Filatov, V. O.; Semenets, V. V.The subject matter of the study is the integration and management of information resources of distributed computing corporate systems within the common information space. The goal of the study is to develop scientifically based models for supporting intelligent technologies for the integration and management of information resources of distributed computing systems. In accordance with the goal of the study, the article deals with the following tasks: to develop a formal-logical model of the information space, to study and formalize agent-oriented tasks, to select the model of a software agent, to substantiate the information technology aimed at effective information management in corporate systems.Публікація Named Entity Recognition Problem for Long Entities in English Texts(2021) Shatalov, O.; Ryabova, N.This paper is related to the problem of natural language processing (NLP), namely the named entity recognition (NER). This paper reveals the features of named entities recognition in English texts using deep learning (DL). The peculiarity of the study was the rather long length of the presented named entities: many of them could include a rather large number of words. The main problem was the amount of text that had to be recognized as a single entity. The results of the research are described here show the effectiveness of using deep neural network architectures for the task of recognizing long named entities in texts in English.Публікація New era in image augmentation – neural radiance fields(ХНУРЕ, 2023) Shashko, O.In this article, a new method of image augmentation using Neural Radiance Fields is being considered. NeRF has significant advantages in generating photo-realistic images with high resolution and fidelity, and it can be used for a wide range of applications, including 3D rendering, virtual or augmented reality, and robotics. Despite its computational cost and limitations in some types of scenes, NeRFs are becoming more and more popular in many fields of computer vision, with ongoing research to improve its limitations. Also, this work describes a practical example of how it can preprocess and augment an existing dataset to use it for a specific aimПублікація On the approach to searching for functional dependences of data in relational systems(ХНУРЕ, 2018) Filatov, V.; Doskalenko, S.The subject matter of the study is information systems built on the basis of relational databases. The goal of the article is to develop a method for re-engineering relational databases that takes into account implicit interrelated functionally dependent data that affect the structure of the logical model. The following results are obtained: the approach to identify previously unknown functional dependencies based on the analysis of a set of relational database data is suggested; the classes of tasks of reengineering relational databases are specified; the stage of developing the target logic diagram which is common for the problems of adaptation and refactoring was studied; the sub-task of checking if the logic diagram of the relational database corresponds to the third normal form within this stage is considered using the synthesis method; it is shown that the solution of this task involves a number of difficulties, in particular, it is necessary to find a set of functional dependencies that are performed on the current instance of the data of a relational database; the approach for finding a set of functional dependencies from an instance of the data of a relational structure is suggested. The direction of further research can be the support of empty values at the stage of identifying functional dependencies as well as the issues of data transfer without any loss from the initial structure of the database to the target data obtained as a result of applying the methods of re-engineering. Conclusions. The approach is suggested to identify previously unknown functional dependencies which are based on the analysis of a set of relational database data. The first step is to get a set of functional dependencies for each relationship. The similar operation for the universal relation of the target database is performed at the second step. At this step, functional dependencies among the attributes of different relationships, that is the interrelationships among the data that were established during the information system operation, can be identified. The method for determining their information novelty is suggested; this method consists in verifying the membership of the functional dependencies of the universal relation while discovering the union of sets of dependencies of individual relations. A promising direction for further research is the development of methods to implement the technology for verifying if the obtained dependencies correspond to the logical model of the domain.Публікація Ontology-based International Degree Recognition(IEEE Computer Society, 2005) Terziyan, V.; Kaykova, O.; Vitko, O.; Titova, L.Bologna process and “eurointegration” needs the development of new opened educational standards (there exist substantial differences between Ukrainian and EU standards in higher education ) none (even the best) of the European universities can provide students with the optimal set of courses of the best experts Educational Ontologies. Semantic Web is not only a technology as many used to name it; Semantic Web is not only an environment as many naming it now; Semantic Web it is a new context within which one should rethink and re- interpret his existing businesses, resources, services, technologies, processes, environments, products etc. to raise them to totally new level of performance. Management and Internationalization of Education are specific fields, in which advances of Semantic Web approach can bring essential effect.Публікація Personalized Adaptation of Learning Environments(2019) Filatov, V.; Yerokhin, A.; Zolotukhin, O.; Kudryavtseva, M.This work is devoted to the development of personalized training systems. A major problem in learning environmens is applying the same approach to all students: teaching materials, time for their mastering, and a training program that is designed in the same way for everyone. Although, each student is individual has his own skills, ability to assimilate the material, his preferences and other. Recently, recommendation systems, of which the system of personalized learning is a part, have become widespread in the learning environment. On the one hand, this shift is due to mathematical approaches, such as machine learning and data mining, that are used in such systems while, on the other hand, the requirements of technological standards "validated" by the World Wide Web Consortium (W3C). According to this symbiosis of mathematical methods and advanced technologies, it is possible to implement a system that has several advantages: identifying current skill levels, building individual learning trajectories, tracking progress, and recommending relevant learning material. The conducted research demonstrates how to make learning environment more adaptive to the users according to their knowledge base, behavior, preferences, and abilities. In this research, a model of a learning ecosystem based on the knowledge and skills annotations is presented. This model is a general model of the all life learning. Second, this thesis focuses on the creation of tools for personalized assessment, recommendation, and advising.Публікація Reengineering relational database on analysis functional dependent attribute(Proceedings of the X Intern. Scient. and Techn. Conf. «Computer Science & Information Technologies» (CSIT’2015), 14-17 sept. 2015. – Lviv, Ukraine. – P. 85-88., 2015) Filatov, V. O.; Radchenko, V.The task of re-engineering information system which is based uses a relational database. An approach to the definition of functionally dependent attributes of the database in step reengineering and modified synthesis algorithm logic of a relational database.Публікація Relational vs non-relational databases(«European Scientific Platform», 2022) Sliusarenko, T.; Filatov, V.; Слюсаренко, Т.; Філатов, В.In this paper the difference between relational and non-relational databases are given.Публікація Report on TRUST project impact on national level(2015) Golovianko, M.Impact of TRUST project results on national level is reported