The International Conference on Security, Fault Tolerance, Intelligence <p>The conference is organized by KPI named after Igor Sikorskyi and is dedicated to the memory of the outstanding scientist, teacher of the Department of Computing, Professor Shirochyn. Co-organizers of the conference on behalf of the university are the Faculty of Informatics and Computer Engineering, the Faculty of Applied Mathematics. The purpose of the conference is to discuss scientific results and exchange experience between scientists conducting research in the field of computer engineering, software engineering and technical education.</p> en-US The International Conference on Security, Fault Tolerance, Intelligence ALGORITHMIC AND METHODOLOGICAL ASPECTS OF THE ORGANIZATION OF THE SOFTWARE SYSTEM OF SUPPORT OF EDUCATIONAL PROCESS PLANNING <p>The article considers aspects of building an information system to support the educational process in higher education. The main problems of automated planning are highlighted and ways and means of their solution are offered.</p> <p><strong>Key words</strong>: automated planning, information system, educational process, artificial intelligence.</p> <p>Fig.: 2. Bibl.: 6.</p> Victor Porev Copyright (c) 2023 2023-11-08 2023-11-08 61 64 PATTERN ONDRAW–ONTOUCH AND ITS USABILITY FOR OWNERDRAW GUI <p>The article considers some aspects of building graphical interfaces with non-standard<br>elements. A pattern for developing applications for event-driven operating platforms is<br>proposed.<br>Key words: GUI, pattern.<br>Fig.: 3. Bibl.: 6.</p> Victor Porev Copyright (c) 2023 2023-11-08 2023-11-08 65 68 SOLVING POISSON EQUATION WITH CONVOLUTIONAL NEURAL NETWORKS <p>РОЗВ’ЯЗУВАННЯ РІВНЯННЯ ПУАССОНА З ЗАСТОСУВАННЯМ ЗГОРТКОВИХ НЕЙРОННИХ МЕРЕЖ</p> Kuzmych Valentyn Novotarskyi Mykhailo Copyright (c) 2023 2023-11-08 2023-11-08 69 76 DEEP IMAGE COMPRESSION SYSTEM FOR THE BETTER RATE-DISTORTION PERFORMANCE <p>This paper analyses the deep image compression system that contains encoder,<br>quantizer, entropy model and decoder optimized by joint rate-distortion framework.<br>Current model implements a channel-level variable quantization network to dynamically<br>allocate and withdraw the bitrates from significant and negligible channels. Its main<br>specific is usage of the variable quantization controller that consists of such components:<br>channel importance module that dynamically learns the importance of channels during the<br>training, and splitting-merging module, which allocates different bitrates of the channels.<br>Quantizer implements the Gaussian mixture model manner. The paper is a continuation of<br>several similar research done before that first provided an idea and architecture of the<br>model. The main goals of the proposed work are to do deeper analysis of the system, verify<br>and improve the model effectiveness. The experiments validate that hyper-parameter<br>tuning approach proposed here successfully improves the rate-distortion performance of<br>image compression in terms of various quality metrics such as PSNR, MS-SSIM and BPP.<br>Key words: convolutional neural networks, Deep learning, Deep Image Compression,<br>Channel-Level Variable Quantization, Deep Neural Networks.<br>Fig.: 3, Tabl.: 3, Bibl.: 11.</p> Vasylenko Dmytro Yevgeniyovych Gordienko Yuri Hryhorovych Stirenko Sergii Hryhorovych Oleksii Aleshchenko Copyright (c) 2023 2023-11-08 2023-11-08 77 85 PREDICTION OF INTERNET ORDERS REJECTIONS BY CUSTOMERS USING THE APPROACHES OF MACHINE LEARNING <p>In the paper the possibility of using machine learning approaches and algorithms to predict rejections by the end user during the Internet order is considered. The two most popular dependency detection algorithms have been studied, namely logistic regression and XGBoost. According to the testing on a large selection of real data, the conclusions were made.</p> <p>Keywords: in-depth data analysis, machine learning, neural networks.</p> Klimenkova Nina Suprun Oleh Лаврій Софія Петрівна Copyright (c) 2023 2023-11-08 2023-11-08 86 89 GENERATIVE ADVERSARIAL NETWORKS FOR IMAGE SUPER RESOLUTION <p>This article demonstrates an experiment with changing one of the functions of the final image processing, the quality of which was improved using deep neural networks. It also describes previously developed methods for training and using neural networks to obtain Super-Resolution images, which have been very successful. Quantitative and qualitative assessment of the training results allows us to conclude that image processing is improved even with a very limited amount of training, and with a larger training the result can be improved even more. Within the specific goal (to get better performance or low resource consumption) and under conditions of full training, this method of image processing can be optimized for video processing. This creates the preconditions for obtaining a video quality improvement system on edge computing devices (gadgets, smartphones, tablets, etc.) and with limited computing resources and reducing the load on the network infrastructure when integrating with online video services.</p> <p><strong>Key words:</strong> convolutional neural networks, deep learning, generative adversarial networks, High-resolution, Low resolution, Super-resolution.</p> Prasolov Andrii Arturovych Gordienko Yuri Hryhorovych Stirenko Sergii Hryhorovych Copyright (c) 2023 2023-11-08 2023-11-08 90 103 HYBRID DEEP NEURAL NETWORKS WITH QUANTUM CIRCUIT FOR IMAGE CLASSIFICATION <p>The article describes hybrid convolutional neural network that uses one quantum circuit for image classification. Described the structure of the network, structure of the quantum circuit, compared several different quantum circuits with different number of qubits. The hybrid neural network was evaluated on MNIST Fashion dataset, which is radically different from MNIST dataset. Performance of hybrid neural network was compared for multiclass classification on MNIST and MNIST Fashion datasets for 4, 6, 8, 10 classes using quantum circuits with 2, 3, 4 qubits.</p> <p><strong>Key words:</strong> neural networks, quantum computing, quantum machine learning, deep learning</p> <p>Fig.: 6. Tabl.: 2. Bibl.: 8.</p> Trochun Yevhenii Volodymyrovych Gordienko Yuri Hryhorovych Stirenko Sergii Hryhorovych Copyright (c) 2023 2023-11-08 2023-11-08 104 109 METHOD OF BUILDING INTELLECTUAL MOBILE SYSTEMS <p>This work describes a method of building intelligent mobile systems. It may be interesting both for the design of a fully independent mobile, partially mobile or stationary system using intelligence, and as part of another intelligent system. The method is based on a set of software using computer vision algorithms and hardware to support the operation of the software and to communicate with the environment.</p> <p><strong>Key words: </strong>computer vision, artificial intelligence, intelligent mobile system.</p> Hlushenok Valerii Korochkin Aleksandr Copyright (c) 2023 2023-11-08 2023-11-08 110 114 TRANSFER LEARNING PROBLEMS AND THEIR POSSIBLE SOLUTIONS <p>The article demonstrates the consideration of image recognition using machine learning. Image recognition and classification is done using Transfer learning, which uses the logic of knowledge transfer from the source domain to the target. Models train on a set of images that are publicly available on the Internet.</p> <p>Fig.: 1.</p> Volokyta Artem Bohdan Mosiyuk Copyright (c) 2023 2023-11-08 2023-11-08 115 117 METHOD FOR GENERATION OF ENGLISH GRAMMAR EXERCISES <p>У статті розглядається метод для генерації граматичних завдань для вивчення англійської мови. Розроблений метод дозволяє створювати граматичні завдання на основі речень з різних джерел за допомогою обробки природної мови (NLP).</p> <p><strong>Ключові слова: </strong>Генерація граматичних завдань, NLP, spaCy.</p> <p>Рис.: 4. Табл.: 0. Бібл.: 3.</p> Lytvyniuk Dmytro Dolgolenko Oleksandr Copyright (c) 2023 2023-11-08 2023-11-08 118 121 CRYPTOGRAPHICALLY STRICT USER IDENTIFICATION BASED ON HASH TRANSFORMATIONS WITH PROGRAMMED COLLISIONS The article proposes an approach to high-speed remote users identification based on hash transformation with programmed collision whose fulfils the role of session paroles. The hash transformation is constructed in the form of a multilayer structure, the elements of which are tabular nonlinear Boolean transformations and permutations. A hash transformation synthesis procedure has been developed. Keywords: remote user entification, hash transfrmation, zero-knowledge identification. Fig .: 2. Tabl .: 0. Bibl .: 11. Daiko Ihor Volodumirivich Markovskyi Oleksandr Petrovich Steblevets Tetyana Oleksandrivna Copyright (c) 2023 2023-11-08 2023-11-08 5 11 EXPERT SYSTEM FOR DIAGNOSING GENETIC DISORDERS The article considers an expert system for diagnosing genetic abnormalities to increase the accuracy of diagnosis by a geneticist. The structure, the scheme of interaction of components and a way of representation of data of expert system are described. Key words: expert system, knowledge base, data representation Fig.: 2. Tabl.: 0. Bibl.: 7. Chyzh Yuliia Simonenko Valeriy Copyright (c) 2023 2023-11-08 2023-11-08 12 15 FAST EXPONENTIATION IN GALOIS FIELDS USING PRECOMPUTATION The article proposes an approach to speed up the calculation of the exponent in Galois fields - the basic operation of a wide range of cryptographic algorithms. The properties of the polynomial square and the Montgomery group reduction with precomputations are used as the main reserve for improving the performance. A mathematical justification is given, a procedure for fast squaring on Galois fields is described. The presentation is illustrated with a numerical example. It has been shown that the proposed approach makes it possible peed up computations up to 3 times. Keywords: exponentiation in Galois field, Montgomery reduction, precomputations. Fig .: 0. Tabl .: 2. Bibl .: 12. Rusanova Olga Veniaminovna Kot Olga Sergeevna Al-Mrayat Ghassan Abdel Jalil Halil Copyright (c) 2023 2023-11-08 2023-11-08 16 23 AUDIT RESULTS INTERPRETATION SYSTEM The article considers the system of audit results interpretation to improve quality and reduce time needed for conducting software audit. The structure, components of the system, their relationships and interactions and a way the system works are described. Key words: software audit, results interpretation, data analysis Fig.: 1. Tabl.: 0. Bibl.: 8. Yevhenii Trochun Dolgolenko Aleksandr Copyright (c) 2023 2023-11-08 2023-11-08 24 28 СИСТЕМА ДЛЯ ПРОВЕДЕННЯ ЕКСКУРСІЙ У ВІРТУАЛЬНІЙ ТА ДОПОВНЕНІЙ РЕАЛЬНОСТІ З ВИКОРИСТАННЯМ ТЕХНОЛОГІЙ КОМП’ЮТЕРНОГО ЗОРУ The paper deals with a system for conducting tours in virtual and augmented reality using computer vision technology. The system is able to recognize known landmarks in the image using a neural network. The model trains on the Google Landmark Dataset. Keywords: augmented reality, neural networks, landmarks, ArcFace, ResNet. Bibl.: 5, fig.: 2. Kravets Anastasiia Simonenko Valeriy Copyright (c) 2023 2023-11-08 2023-11-08 29 31 AUDITING UNIX SYSTEMS The article considers the Linux audit system. It is an outstanding way for sysadmins to create a log rule for nearly every action on a data center server. Using this system means you can track events, record the events, and even detect abuse or unauthorized activity, via the log files.. Key words: Unix system, auditing programs, super-user, events. Simonenko Andrey Copyright (c) 2023 2023-11-08 2023-11-08 32 35 МАСШТАБОВАНА ІНФОРМАЦІЙНА СИСТЕМА УПРАВЛІННЯ НАВЧАЛЬНИМ ЗАКЛАДОМ The article proposes an approach to creating an architecture of a scalable Management Information System (MIS) of educational institutions. A set of components and functions for building an information system based on the proposed approach is proposed. Keywords: Management Information System, Educational Management Information System, data, information, authentication, authorization. Maryna Sivachenko Iryna Klymenko Copyright (c) 2023 2023-11-08 2023-11-08 36 40 RATINGS CALCULATIONS MEANS OF THE DETECTED ROUTES BASED ON FUZZY LOGIC METHOD The article presents a method of calculating the routes ratings in the software-defined transport network, found with a modified inverse wave algorithm, using a fuzzy inference system. The effectiveness of this method is the ability to take into account many parameters simultaneously for decision making without performing complex mathematical calculations. Also, this method helps to reduce the likelihood of re-routing the path during routing. Keywords: route ratings, SDN controller, modified inverse wave algorithm, fuzzy inference system, re-routing. Fig.: 5. Tabl.: 2. Bibl.: 5. Yuliia Hrabovenko Yurii Kulakov Copyright (c) 2023 2023-11-08 2023-11-08 41 46 MULTIPATH ROUTING METHOD IN CLUSTER-BASED VEHICULAR NETWORKS The paper considers the issues of traffic engineering (TE) in cluster-based vehicular networks. The expediency of using multipath routing in TE is given. A modified cluster-based algorithm for searching multiple paths in a software-defined networks taking into account the parameters of quality of service (QoS) is proposed. Key words: Multipath Routing, cluster-based vehicular networks, Software-Defined Networking (SDN), Quality of Service (QoS). Fig.: 4. Tabl.: 1. Bibl.: 6. Yurii Kulakov Iryna Hrabovenko Copyright (c) 2023 2023-11-08 2023-11-08 47 52 LOAD BALANCING ISSUES IN CLUSTER GRID SYSTEMS BASED ON INPUT INFORMATION The paper deals with the issues of load balancing in cluster grid systems. To do this, the method of assignment is used, which takes into account the presence of input data tasks in certain computing nodes. Performance testing is performed on a set of tasks, a certain percentage of which have common input data. Key words: clusters, grid systems, distributed computing, load balancing. Fig.: 3. Bibl.: 3. Simonenko Valery Pavlovich Vitaliy Semenchenko Copyright (c) 2023 2023-11-08 2023-11-08 53 55 ISSUES AND METHODS OF BIOMETRIC AUTHENTICATION <p>Abstract. This paper dials with biometric authentication methods, such as fingerprint recognition, face recognition, their advantages and disadvantages and additional biometry such as behavior keystroke-based authentication method, taking in consideration existing solution. Also own developed solution based on neural network learning is presented and described.<br />Key words: biometrics, authentication, user behavior, keystroke, neural networks.<br />Fig.: 3. Tabl.: 1. Bibl.: 4.</p> Kosareva Anastasiia Rehida Pavlo Copyright (c) 2023 2023-11-08 2023-11-08 56 60