Видання НТУ "ХПІ"
Постійне посилання на розділhttps://repository.kpi.kharkov.ua/handle/KhPI-Press/62886
Переглянути
10 результатів
Результати пошуку
Документ Blind signal separation applications and methods(Національний технічний університет "Харківський політехнічний інститут", 2024) Monastyrskyi, MykytaBlind signal separation is the task of separating the given mixture signal into two or more corresponding sources. It finds an application in many fields of human activity such as medicine, telecommunications, art and many more and is a crucial task in signal processing. However, the task itself appears to be quite challenging due to its ill-posed nature. Despite that many modern machine learning-based approaches achieve the state-of-the-art results in different blind source separation tasks (e. g. audio or music source separation) however these methods can suffer from unwanted artifacts in the source signals estimates. This paper presents an overview of the methods for blind source separation covering methods from traditional statistical ones to modern machine learning-based approaches and applications of the results of blind source separation task. Moreover, we discuss some potential areas of research in the field of blind source separation to facilitate further research and develop powerful solutions for this task.Документ Sequential intrusion detection system for zero-trust cyber defense of IoT/IIoT networks(Національний технічний університет "Харківський політехнічний інститут", 2024) Sobchuk, Valentyn; Pykhnivskyi, Roman; Barabash, Oleg; Korotin, Serhii; Omarov, ShakhinThe Internet of Things (IoT) and the Industrial Internet of Things (IIoT) and their widespread application make them attractive targets for cyber attacks. Traditional cybersecurity methods such as firewalls and antivirus software are not always effective in protecting IoT/IIoT networks due to their heterogeneity and large number of connected devices. The zero-trust principle can be more effective in protecting IoT/IIoT networks. This principle assumes on no inherent trustworthiness of any user, device, or traffic, requiring authorization and verification before accessing any network resource. This article presents a zero-trust-based intrusion detection system (IDS) that uses machine learning to secure IoT/IIoT networks. The aim of this article is to develop a two-component IDS for detecting and classifying cyber-attacks. The proposed design for an Intrusion Detection System (IDS) achieves high accuracy in detecting attacks while maintaining optimal performance and minimizing additional computational costs. This is especially crucial for real-time network monitoring in IoT/IIoT environments.Документ Intrusion detection model based on improved transformer(Національний технічний університет "Харківський політехнічний інститут", 2024) Gavrylenko, Svitlana; Poltoratskyi, Vadym; Nechyporenko, AlinaThe object of the study is the process of identifying the state of a computer network. The subject of the study are the methods of identifying the state of computer networks. The purpose of the paper is to improve the efficacy of intrusion detection in computer networks by developing a method based on transformer models. The results obtained. The work analyzes traditional machine learning algorithms, deep learning methods and considers the advantages of using transformer models. A method for detecting intrusions in computer networks is proposed. This method differs from known approaches by utilizing the Vision Transformer for Small-size Datasets (ViTSD) deep learning algorithm. The method incorporates procedures to reduce the correlation of input data and transform data into a specific format required for model operations. The developed methods are implemented using Python and the GOOGLE COLAB cloud service with Jupyter Notebook. Conclusions. Experiments confirmed the efficiency of the proposed method. The use of the developed method based on the ViTSD algorithm and the data preprocessing procedure increases the model's accuracy to 98.7%. This makes it possible to recommend it for practical use, in order to improve the accuracy of identifying the state of a computer system.Документ Performance comparison of U-Net and LinkNet with different encoders for reforestation detection(Національний технічний університет "Харківський політехнічний інститут", 2024) Podorozhniak, Andrii; Onishchenko, Daniil; Liubchenko, Nataliia; Grynov, DenysThe subject of study is analysis of performance of artificial intelligence systems with different architectures for reforestation detection. The goal is to implement, train and evaluate system with different models for deforestation and reforestation detection. The tasks are to study problems and potential solutions in forestry for reforestation detection and present own solution. As part of model comparison, results are presented for different artificial neural network architectures with different encoders. For training and testing purpose custom dataset was created, which includes different areas of territory of Ukraine within different timestamps. Main research methods are literature analysis, experiment and case study. As a result of analysis of modern artificial intelligence methods, machine learning, deep learning and convolutional neural networks, high-precision algorithms U-Net and LinkNet were chosen for system implementation. Conclusions. The studied problem was stated formally and broken down in smaller steps; possible solutions were studied and proposed solution was described in details. Necessary mathematical background for analysis of the performance was provided. As part of the development, accurate deforestation/reforestation module was created. All analysis results were listed and a comparison of the studied algorithms was presented.Документ Performance evaluation of python libraries for multithreading data processing(Національний технічний університет "Харківський політехнічний інститут", 2024) Krivtsov, Serhii; Parfeniuk, Yurii; Bazilevych, Kseniia; Meniailov, Ievgen; Chumachenko, DmytroTopicality. The rapid growth of data in various domains has necessitated the development of efficient tools and libraries for data processing and analysis. Python, a popular programming language for data analysis, offers several libraries, such as NumPy and Numba, for numerical computations. However, there is a lack of comprehensive studies comparing the performance of these libraries across different tasks and data sizes. The aim of the study. This study aims to fill this gap by comparing the performance of Python, NumPy, Numba, and Numba.Cuda across different tasks and data sizes. Additionally, it evaluates the impact of multithreading and GPU utilization on computation speed. Research results. The results indicate that Numba and Numba.Cuda significantly optimizes the performance of Python applications, especially for functions involving loops and array operations. Moreover, GPU and multithreading in Python further enhance computation speed, although with certain limitations and considerations. Conclusion. This study contributes to the field by providing valuable insights into the performance of different Python libraries and the effectiveness of GPU and multithreading in Python, thereby aiding researchers and practitioners in selecting the most suitable tools for their computational needs.Документ Maximizing solar photovoltaic system efficiency by multivariate linear regression based maximum power point tracking using machine learning(Національний технічний університет "Харківський політехнічний інститут", 2024) Paquianadin, V.; Navin Sam, K.; Koperundevi, G.Introduction. In recent times, there has been a growing popularity of photovoltaic (PV) systems, primarily due to their numerous advantages in the field of renewable energy. One crucial and challenging task in PV systems is tracking the maximum power point (MPP), which is essential for enhancing their efficiency. Aim. PV systems face two main challenges. Firstly, they exhibit low efficiency in generating electric power, particularly in situations of low irradiation. Secondly, there is a strong connection between the power output of solar arrays and the constantly changing weather conditions. This interdependence can lead to load mismatch, where the maximum power is not effectively extracted and delivered to the load. This problem is commonly referred to as the maximum power point tracking (MPPT) problem various control methods for MPPT have been suggested to optimize the peak power output and overall generation efficiency of PV systems. Methodology. This article presents a novel approach to maximize the efficiency of solar PV systems by tracking the MPP and dynamic response of the system is investigated. Originality. The technique involves a multivariate linear regression (MLR) machine learning algorithm to predict the MPP for any value of irradiance level and temperature, based on data collected from the solar PV generator specifications. This information is then used to calculate the duty ratio for the boost converter. Results. MATLAB/Simulink simulations and experimental results demonstrate that this approach consistently achieves a mean efficiency of over 96 % in the steady-state operation of the PV system, even under variable irradiance level and temperature. Practical value. The improved efficiency of 96 % of the proposed MLR based MPP in the steady-state operation extracting maximum from PV system, adds more value. The same is evidently proved by the hardware results. References 24, table 4, figures 14.Документ Application of heterogeneous ensembles in problems of computer system state identification(Національний технічний університет "Харківський політехнічний інститут", 2023) Hornostal, Oleksii; Gavrylenko, SvitlanaThe object of the study is the process of identifying anomalies in the operation of a computer system (CS). The subject of the study is ensemble methods for identifying the state of the CS. The goal of the study is to improve the performance of ensemble classifiers based on heterogeneous models. Methods used: machine learning methods, homogeneous and heterogeneous ensemble classifiers, Pasting and Bootstrapping technologies. Results obtained: a comparative analysis of the use of homogeneous and heterogeneous bagging ensembles in data classification problems was carried out. The effectiveness of various approaches to the selection of base ensemble classifiers has been studied. A method for identifying the state of a computer system, based on the heterogeneous bagging ensemble was proposed. Experimental studies made it possible to confirm the main theoretical assumptions, as well as evaluate the efficiency of the constructed heterogeneous ensembles. Conclusions. Based on the results of the study, the method for constructing a heterogeneous bagging ensemble classifier, which differs from known methods in the procedure for selecting base models was proposed. It made possible to increase the classification accuracy. Further development of this research could include the creating and integration of dissimilarity metrics as well as other quantitative metrics for a more accurate and balanced base model selection procedure, which would further improve the performance of the computer system state classifier.Документ Enhancing power system security using soft computing and machine learning(Національний технічний університет "Харківський політехнічний інститут", 2023) Venkatesh, Peruthambi; Visali, NagalamadakaTo guarantee proper operation of the system, the suggested method infers the loss of a single transmission line in order to calculate a contingency rating. Methods. The proposed mathematical model with the machine learning with particle swarm optimization algorithm has been used to observe the stability analysis with and without the unified power flow controller and interline power flow controller, as well as the associated costs. This allows for rapid prediction of the most affected transmission line and the location for compensation. Results. Many contingency conditions, such as the failure of a single transmission line and change in the load, are built into the power system. The single transmission line outage and load fluctuation used to determine the contingency ranking are the primary emphasis of this work. Practical value. In order to set up a safe transmission power system, the suggested stability analysis has been quite helpful.Документ Метод оцінки та підвищення користувальницького досвіду абонентів в програмно-конфігурованих мережах на основі використання машинного навчання(Національний технічний університет "Харківський політехнічний інститут", 2023) Аль-Мудхафар, Акіл Абдулхуссейн М.; Смірнова, Тетяна Віталіївна; Буравченко, Костянтин Олегович; Смірнов, Олексій АнатолійовичЕволюційні процеси, які в першу чергу торкнулися комп'ютерних технологій, призвели до появи кількох типів обчислювальних мереж, що представляють сукупність комп'ютерних пристроїв, об'єднаних в одну систему. Основним призначенням такої системи є доступ користувачів до спільних ресурсів та можливість обміну даними між абонентами у процесі роботи. Такі мережі називаються програмно-конфігурованими – SDN. Мережі SDN вже давно стали основою побудови телекомунікаційних мереж операторського класу. Проте, в них є певна кількість недоліків, які необхідно усунути. Об’єктом дослідження є процес оцінки та підвищення користувальницького досвіду абонентів в програмно-конфігурованих мережах. Предметом дослідження є метод оцінки та підвищення користувальницького досвіду абонентів в програмно-конфігурованих мережах на основі використання машинного навчання. Мета роботи полягає у розробці моделі та відповідного методу оцінки якості користувальницького досвіду абонентів мереж SDN. У результаті дослідження вперше було розроблено метод оцінки та підвищення користувальницького досвіду абонентів мереж SDN на основі використання машинного навчання. Метод полягає у послідовному проведенні автоматизованого опитування користувачів, вимірюванні показників якості обслуговування абонентів, виборі й побудові регресійної моделі із множини визначених моделей та керування користувальницьким досвідом за виміряними параметрами якості обслуговування абонентів мережі SDN. Розроблений метод на відміну від відомих, надає змогу підвищувати якість користувальницького досвіду у режимі реального часу. Висновки. Проведене дослідження існуючих механізмів керування користувальницьким досвідом абонентів та аналіз регресійних моделей на можливість їх використання для встановлення взаємозв’язку між параметрами мережі та користувальницьким досвідом, дозволило розробити узагальнену модель оцінки та підвищення користувальницького досвіду абонентів мереж SDN, на основі використання машинного навчання, та розробити алгоритм роботи методу. Розроблений метод дозоляє будувати точні моделі взаємозв’язку параметрів QoE та QoS та підвищує на величину до 10% якість користувальницького досвіду абонентів мереж SDN.Документ Usage of Mask R-CNN for automatic license plate recognition(Національний технічний університет "Харківський політехнічний інститут", 2023) Podorozhniak, Andrii; Liubchenko, Nataliia; Sobol, Maksym; Onishchenko, DaniilThe subject of study is the creation process of an artificial intelligence system for automatic license plate detection. The goal is to achieve high license plate recognition accuracy on large camera angles with character extraction. The tasks are to study existing license plate recognition technics and to create an artificial intelligence system that works on big shooting camera angles with the help of modern machine learning solution – deep learning. As part of the research, both hardware and software-based solutions were studied and developed. For testing purposes, different datasets and competing systems were used. Main research methods are experiment, literature analysis and case study for hardware systems As a result of analysis of modern methods, Mask R-CNN algorithm was chosen due to high accuracy. Conclusions. Problem statement was declared; solution methods were listed and characterized; main algorithm was chosen and mathematical background was presented. As part of the development procedure, accurate automatic license plate system was presented and implemented in different hardware environments. Comparison of the network with existing competitive systems was made.Different object detection characteristics, such as Recall, Precision and F1-Score, were calculated. The acquired results show that developed system on Mask R-CNN algorithm process images with high accuracy on large camera shooting angles.