Радюк Павло Михайлович

Постійний URI для цього зібрання

https://hdl.handle.net/11300/26301

Переглянути

Зараз відображається 1 - 20 з 20

A Novel Feature Vector for ECG Classification using Deep Learning
(2023) Радюк, Павло Михайлович; Radiuk, Pavlo M.; Kovalchuk, Oleksii; Oleksander, Barmak; Petrovskyi, Sergіi; Krak, Iurii
In the past decade, deep learning techniques have been widely used in the healthcare industry to detect heartbeats and diagnose heart conditions. However, these tools have been criticized for being a “black box” and lacking transparency. Therefore, in this paper, we propose a new approach to making the classification results obtained by deep learning more comprehensible. We suggest forming a vector of features based on ECG signals that correspond to specific heart conditions. This vector includes measurable characteristics of the cardiac cycle, such as wave durations and amplitudes, which are typical and understandable to healthcare professionals. This feature vector serves as input data for a deep neural network that acts as a feature encoder and classifier. Our computational experiments with the handcrafted feature vector achieved an average accuracy of 98.69%, comparable to other deep learning tools based on the complete cardiac cycle. The results of this study suggest that future research should focus on developing interpretable deep learning tools that are transparent and comprehensible to healthcare professionals.
Human-in-the-Loop Approach Based on MRI and ECG for Healthcare Diagnosis
(2022) Радюк, Павло Михайлович; Radiuk, Pavlo M.; Kovalchuk, Oleksii; Slobodzian, Vitalii; Oleksander, Barmak; Krak, Iurii; Manziuka, Eduard
The presented study investigates a human-centric approach to implementing human-intheloop models for healthcare diagnostics. The following tasks were considered and addressed in this work: a) identify the features necessary for future healthcare diagnosis based on electrocardiogram signals in the human-in-the-loop model: P, T-peaks, QRScomplex, PQ and ST segments, and b) detect inflammatory processes in the heart muscle (myocardium) based on cardiac magnetic resonance imaging. As a result of our investigation, a novel approach was proposed for embedding (integrating) clinical knowledge about the nature of these phenomena into the electrocardiogram signal and magnetic resonance imaging. Domain knowledge about the sample’s nature is encoded similarly to the input information. Moreover, the convolution operation within our approach serves as an embedding mechanism. The results presented in the article are a starting point for using the models obtained by the proposed approach (human-in-the-loop models) for classification problems using deep learning and convolutional neural networks. Also, visual analysis shows the proposed approaches’ ability to solve practical clinical problems. It also ensures transparent interpretation of the obtained results as the human-in-the-loop model, which, in turn, is built according to the human-centric approach. Overall, our contribution allows the implementation of a scheme for obtaining artificial intelligence solutions based on the principles of trust in them.
An Ensemble Machine Learning Approach for Twitter Sentiment Analysis
(2022) Радюк, Павло Михайлович; Radiuk, Pavlo M.; Pavlova, Olga; Hrypynska, Nadiia
The presented study addresses the issue of classifying emotional expressions based on small texts (tweets) extracted from the social network Twitter. In this paper, we propose a novel approach to preprocessing tweets to fit them more effectively into the classification model. Moreover, we suggest utilizing two types of features, namely unigrams and bigrams, to expand the feature vector. The classification task of emotional expressions was performed according to several machine learning algorithms: raw random forest, gradient boosting random forest, support vector machine, multilayer perceptron, recurrent neural network, and convolutional neural network. The feature vector elements are presented as sparse and dense subvectors. As a result of computational experiments, it was found that the “appearance” in the reflection of the sparse vector provided higher performance than the “regularity.” The experiments also showed that deep learning approaches performed better than traditional machine learning techniques. Consequently, the best recurrent neural network achieved an accuracy of 83.0% on the test dataset, while the best convolutional neural network reached 83.34%. At the same time, it was discovered that the convolutional model with the support vector machine classifier showed better performance than the single convolutional neural network. Overall, the proposed ensemble method based on receiving the most votes according to the five best models’ predictions has reached an absolute accuracy of 85.71%, proving its practical usefulness.
Method of Facial Geometric Feature Representation for Information Security Systems
(2022) Радюк, Павло Михайлович; Radiuk, Pavlo M.; Kalyta, Oleg; Krak, Iurii; Barmak, Olexander; Wojcik, Waldemar
Throughout human history, emotional manifestations have played a major role in interpersonal interaction among humans in all areas of society. In particular, information security systems for visual surveillance, based on recognizing emotional states by facial expressions, have recently become highly relevant. In this paper, we propose a method of representing geometric facial features, which aims to enhance the functioning of visual surveillance for information security systems. The method is designed to automatically reflect the facial expressions of human emotions in the form of quantitative characteristics of geometric shapes. It uses software-generated landmarks for constructing specific geometric characteristics of the face, which serve as input data for the method. Our method consists in forming seven geometric shapes based on predefined landmarks, with the subsequent quantitative expression of these shapes. The method derives quantitative features of seven forms, which are further used to identify emotional facial states. We validated the proposed method using hyperplane classification and compared its performance with analogs. As such, the classification model, which was constructed based on the proposed method, achieved a classification accuracy of 92.73% and slightly surpassed the analogs in other statistical indicators. Overall, the results of computational experiments confirmed the effectiveness of the proposed method for identifying changes in a person’s emotional state by facial expressions. In addition, the use of simple mathematical calculations in our method has significantly reduced the computational complexity against analogs.
Convolutional Neural Network for Parking Slots Detection
(2022) Radiuk, Pavlo M.; Радюк, Павло Михайлович; Pavlova, Olga; Houda, El Bouhissib; Avsiyevych, Volodymyr; Kovalenko, Volodymyr
With the rapid growth of transport number on our streets, the need for finding a vacant parking spot today could most of the time be problematic, but even more in the coming future. Smart parking solutions have proved their usefulness for the localization of unoccupied parking spots. Nowadays, surveillance cameras can provide more advanced solutions for smart cities by finding vacant parking spots and providing cars safety in the public parking area. Based on the analysis, Google Cloud Vision technology has been selected to develop a cyber-physical system for smart parking based on computer vision technology. Moreover, a new model based on the fine-tuned convolutional neural network has been developed to detect empty and occupied slots in the parking lot images collected from the KhNUParking dataset. Based on the achieved results, the performance of parking lots’ detections can be simplified, and its accuracy improved. The Google Cloud Vision technology as parking slots detector and a pre-trained convolutional neural network as a feature extractor and a classifier were selected to develop a cyber-physical system for smart parking. As a result of the computational investigation, the proposed fine-tuned CNN managed to process 66 parking slots in roughly 0.14 seconds on a single GPU with an accuracy of 85.4%, demonstrating decent performance and practical value. Overall, all considered approaches contain strengths and weaknesses and might be applied to the task of parking slots detection depending on the number of images, CCTV angle, and weather conditions.
Detection of Early Pneumonia on Individual CT Scans with Dilated Convolutions
(2021) Радюк, Павло Михайлович; Radiuk, P. M.; Krak, Iurii; Barmak, Olexander
Over the past decades, pneumonia has been considered one of the most dangerous diseases, leading to severe consequences in a short time. Without proper and timely treatment, pneumonia can lead to fatal consequences. Thus, early diagnosis and detection of this lung disease are crucial in successful treatment and constant monitoring. Indeed, there is a high demand for the development of medical image technologies for disease identification. In this paper, we propose a novel information technology for robust feature identification and early detection of pneumonia on computer tomography scans. We also propose a new modified convolutional neural network as a core feature extractor. An effective dilated convolution operation with different rates, combining features of various receptive fields, was utilized to detect and analyze visual deviations in targeted images. Due to applying the dilated convolutions, the network avoids significant losses of objects' spatial information while providing low computational losses. The investigated model classifies computed tomography images with a validation accuracy of up to 96.12%. Overall, our approach requires much fewer computing resources, proving its effectiveness for solving practical problems on available computing devices.
Information Technology for Early Diagnosis of Pneumonia on Individual Radiographs
(2020) Радюк, Павло Михайлович; Krak, Iurii; Oleksander, Barmak; Radiuk, Pavlo M.
Nowadays, pneumonia remains a disease with one of the highest death rates around the world. The ailment’s pathogen instantly causes a large amount of fluid into the lungs, leading to acute exacerbation. Without preliminary examination and timely treatment, pneumonia can result in severe pulmonary complications. Consequently, early diagnosis of pneumonia becomes a decisive factor in treatment and monitoring the disease. Therefore, information systems that can identify early pneumonia on the Chest X-ray images are becoming more demanding nowadays. An individual approach to a person might be a promising way of early diagnosis. The presented study considers an approach to feature extraction of the early stage of pneumonia and identifying the disease using a relatively simple convolutional neural network. With only three convolutional and two linearization layers, the proposed architecture classifies radiographs with 90.87% accuracy, approaching the results of deep multilayer and resource-intensive architectures in classification accuracy and exceeding them in time efficiency. Our approach requires relatively fewer computing resources, confirming its efficiency in solving practical tasks on available computing devices.
Heuristic Architecture Search Using Network Morphism for Chest X-Ray Classification
(2020) Радюк, Павло Михайлович; Radiuk, Pavlo M.; Kutucu, Hakan
Nowadays, the demand for medical image computing is exceptionally high. This growth was mostly driven by the manual development of machine learning models, in particular neural networks. However, due to the constant evolution of domain requirements, manual model development has become insufficient. The present study proposes a heuristic architecture search that can be in an excellent service for the task of medical image classification. We implemented a novel approach called network morphism to the search algorithm. The proposed search method utilizes the enforced hill-climbing algorithm and functional-saving modifications. As a result of computational experiments, the search method found the optimal architecture in 28 GPU hours. The model formed by the found architecture achieved performance of 73.2% in validation accuracy and 84.5% in AUC on the validation dataset that is competitive to the state-of-the-art hand-crafted networks. Moreover, the proposed search method managed to find the architecture that contains four times fewer parameters. Besides, the model requires almost ten times less physical memory, which may indicate the practical usefulness of our method in medical image analysis.
Методи інтелектуального аналізу даних в інформаційних технологіях : методичні вказівки до самостійної роботи студентів спеціальностей 121 «Інженерія програмного забезпечення» та 113 «Прикладна математика»
(Хмельницький : ХНУ, 2019) Грипинська, Н. В.; Радюк, Павло Михайлович
Facial Emotion Recognition for Photo and Video Surveillance Based on Machine Learning and Visual Analytics
(2023) Kalyta, Oleg; Barmak, Olexander; Radiuk, Pavlo M.; Радюк, Павло Михайлович; Krak, Iurii
Modern video surveillance systems mainly rely on human operators to monitor and interpret the behavior of individuals in real time, which may lead to severe delays in responding to an emergency. Therefore, there is a need for continued research into the designing of interpretable and more transparent emotion recognition models that can effectively detect emotions in safety video surveillance systems. This study proposes a novel technique incorporating a straightforward model for detecting sudden changes in a person’s emotional state using low-resolution photos and video frames from surveillance cameras. The proposed technique includes a method of the geometric interpretation of facial areas to extract features of facial expression, the method of hyperplane classiﬁcation for identifying emotional states in the feature vector space, and the principles of visual analytics and “human in the loop” to obtain transparent and interpretable classiﬁers. The experimental testing using the developed software prototype validates the scientiﬁc claims of the proposed technique. Its implementation improves the reliability of abnormal behavior detection via facial expressions by 0.91–2.20%, depending on different emotions and environmental conditions. Moreover, it decreases the error probability in identifying sudden emotional shifts by 0.23–2.21% compared to existing counterparts. Future research will aim to improve the approach quantitatively and address the limitations discussed in this paper.
Intelligent data analysis using artificial neural networks for decision making in the education domain
(2021) Radiuk, Pavlo M.; Радюк, Павло Михайлович; Olexander, Mazurets; Мазурець, О. В.; Skrypnyk, Tetiana; Moroz, Oleksandr; Скрипник, Т. К.; Мороз, О. В.
Nowadays, applying educational intelligent data analysis (EIDA) seems relevant for improving the educational process based on big data. It implies developing and improving the methods of processing collected data in educational institutions to understand academic issues better. Over the past decades, artificial neural networks (ANNs) have been recognized as the most prominent techniques for learning analytics. In this work, we systematized the recent scientific literature in EIDA with ANNs. The paper analyzes the applications of ANN to EIDA and discusses the computational issues in the EIDA domain. According to the investigation, most educational data mining tasks are addressed by controlled learning models, such as classification, regression, and time-series prediction. Most in-depth methods used in the EIDA domain are traditional types of ANN. Well-known techniques such as multi-year perceptron and deep long short‐term memory networks have been mainly used for classification and prediction tasks within the education sphere. However, the difficulty of interpreting the results produced by ANNs has also been a challenge for intelligent data practitioners in any domain, including education.
Web-based information technology for classifying and interpreting early pneumonia based on fine-tuned convolutional neural network
(2021) Radiuk, P. M.; Радюк, Павло Михайлович; Radiuk, Pavlo M.; Бармак, О. В.; Бармак, Олександр; Barmak, Olexander
There have been rapid development and application of computer methods and information systems in digital medical diagnosis in recent years. However, although computer methods of medical imaging have proven helpful in diagnosing lung disease, for detecting early pneumonia on chest X-rays, the problem of cooperation between professional radiologists and specialists in computer science remains urgent. Thus, to address this issue, we propose information technology that medical professionals can employ to detect pneumonia on chest X-rays and interpret the results of the digital diagnosis. The technology is presented as a web-oriented system with an available and intuitive user interface. The information system contains three primary components: a module for disease prediction based on a classification model, a module responsible for hyperparameter tuning of the model, and a module for interpreting the diagnosis results. In combination, these three modules form a feasible tool to facilitate medical research in radiology. Moreover, a web-based system with a local server allows storing personal patient data on the user's computing device, as all calculations are performed locally.
Modular search space for automated design of neural architecture
(2020) Radiuk, Pavlo M.; Радюк, Павло Михайлович
The past years of research have shown that automated machine learning and neural architecture search are an inevitable future for image recognition tasks. In addition, a crucial aspect of any automated search is the predefined search space. As many studies have demonstrated, the modularization technique may simplify the underlying search space by fostering successful blocks’ reuse. In this regard, the presented research aims to investigate the use of modularization in automated machine learning. In this paper, we propose and examine a modularized space based on the substantial limitation to seeded building blocks for neural architecture search. To make a search space viable, we presented all modules of the space as multisectoral networks. Therefore, each architecture within the search space could be unequivocally described by a vector. In our case, a module was a predetermined number of parameterized layers with information about their relationships. We applied the proposed modular search space to a genetic algorithm and evaluated it on the CIFAR-10 and CIFAR-100 datasets based on modules from the NAS-Bench-201 benchmark. To address the complexity of the search space, we randomly sampled twenty-five modules and included them in the database. Overall, our approach retrieved competitive architectures in averaged 8 GPU hours. The final model achieved the validation accuracy of 89.1% and 73.2% on the CIFAR-10 and CIFAR- 100 datasets, respectively. The learning process required slightly fewer GPU hours compared to other approaches, and the resulting network contained fewer parameters to signal lightness of the model. Such an outcome may indicate the considerable potential of sophisticated ranking approaches. The conducted experiments also revealed that a straightforward and transparent search space could address the challenging task of neural architecture search. Further research should be undertaken to explore how the predefined knowledge base of modules could benefit modular search space.
Applying 3D U-Net Architecture to the Task of Multi-Organ Segmentation in Computed Tomography
(2020) Radiuk, Pavlo M.; Радюк, Павло Михайлович
The achievement of high-precision segmentation in medical image analysis has been an active direction of research over the past decade. Significant success in medical imaging tasks has been feasible due to the employment of deep learning methods, including convolutional neural networks (CNNs). Convolutional architectures have been mostly applied to homogeneous medical datasets with separate organs. Nevertheless, the segmentation of volumetric medical images of several organs remains an open question. In this paper, we investigate fully convolutional neural networks (FCNs) and propose a modified 3D U-Net architecture devoted to the processing of computed tomography (CT) volumetric images in the automatic semantic segmentation tasks. To benchmark the architecture, we utilised the differentiable Sørensen-Dice similarity coefficient (SDSC) as a validation metric and optimised it on the training data by minimising the loss function. Our hand-crafted architecture was trained and tested on the manually compiled dataset of CT scans. The improved 3D UNet architecture achieved the average SDSC score of 84.8 % on testing subset among multiple abdominal organs. We also compared our architecture with recognised state-of-the-art results and demonstrated that 3D U-Net based architectures could achieve competitive performance and efficiency in the multi-organ segmentation task.
Application of a genetic algorithm to search for the optimal convolutional neural network architecture with weight distribution
(2020) Radiuk, P. M.; Радюк, Павло Михайлович; Radiuk, Pavlo M.; Radiuk, Pavlo M.
In the past decade, a new way in neural networks research called Network architectures search has demonstrated noticeable results in the design of architectures for image segmentation and classification. Despite the considerable success of the architecture search in image segmentation and classification, it is still an unresolved and urgent problem. Moreover, the neural architecture search is also a highly computationally expensive task. This work proposes a new approach based on a genetic algorithm to search for the optimal convolutional neural network architecture. We integrated a genetic algorithm with standard stochastic gradient descent that implements weight distribution across all architecture solutions. This approach utilises a genetic algorithm to design a sub-graph of a convolution cell, which maximises the accuracy on the validation set. We show the performance of our approach on the CIFAR-10 and CIFAR-100 datasets with a final accuracy of 93.21% and 78.89%, respectively. The main scientific contribution of our work is the combination of genetic algorithm with weight distribution in the architecture search tasks that achieve similar to state-of-the-art results on a single GPU.
A Framework for Exploring and Modelling Neural Architecture Search Methods
(Lviv, Ukraine, 2020) Радюк, Павло Михайлович; Radiuk, Pavlo M.; Радюк, П. М.; Hrypynska, N.
For the past years, many researchers and engineers have been developing and optimising deep neural networks (DNN). The process of neural architecture design and tuning its hyperparameters remains monotonous, timeconsuming, and do not always ensure optimal results. In his regard, the auto- matic design of machine learning (AutoML) has been widely utilised, and neural architecture search (NAS) has been actively developing in recent years. De- spite meaningful advances in the field of NAS, a unified, systematic approach to explore and compare search methods has not been established yet. In this paper, we aim to close this knowledge gap by summarising search decisions and strategies and propose a schematic framework. It applies quantitative and qualitative metrics for prototyping, comparing, and benchmarking the NAS methods. Moreover, our framework enables categorising critical areas to search for better neural architectures.
Neuroevolution of convolutional neural networks for the classification of lung cancer images
(2018) Радюк, Павло Михайлович; Radiuk, Pavlo M.
Convolutional neural networks demonstrate impressive results during medical imaging of lung cancer. It may be possible to make diagnoses with convolutional neural networks on conventional chest X-rays that are definitively apparent on subsequently computed tomography and biopsy. Computer vision may reduce the need for further evaluation with invasive testing or prevent errors of missed diagnoses. Using over twelve thousand images of proven lung cancer from the Prostate, Lung, Colorectal, and Ovarian dataset, we developed an algorithm to predict the presence or absence of lung cancer. The classification algorithm has achieved an accuracy of 96.09% with a positive predictive value of 99.11% and a negative predictive value of 93.25%.
Impact of Training Set Batch Size on the Performance of Convolutional Neural Networks for Diverse Datasets
(2017) Радюк, Павло Михайлович; Radiuk, Pavlo M.; Радюк, П. М.
A problem of improving the performance of convolutional neural networks is considered. A parameter of the training set is investigated. The parameter is the batch size. The goal is to find an impact of training set batch size on the performance. To get consistent results, diverse datasets are used They are MNIST and CIFAR-10. Simplicity of the MNIST dataset stands against complexity of the CIFAR-10 dataset, although the simpler dataset has 10 classes as well as the more complicated one To achieve acceptable testing results, various convolutional neural network architectures are selected for the MNIST and CIFAR-10 datasets, with two and five convolutional layers, respectively. The assumption about the dependence of the recognition accuracy on the batch size value is confirmed: the larger the batch size value the higher the recognition accuracy. Another assumption about the impact of the type of the batch size value on the CNN performance is not confirmed.
Інформаційна технологія візуального аналізу рентгенівських зображень для інтерпретації результатів діагностування пневмонії
(2021) Радюк, Павло Михайлович; Радюк, П. М.; Radiuk, Pavlo M.; Бармак, О. В.; Бармак, Олександр
На сьогодні пневмонія є одним із поширеніших та найбільш серйозних легеневих захворювань у всьому світі. Раннє діагностування пневмонії є ключовим чинником її успішного лікування. Для розв’язання актуального завдання в галузі цифрового діагностування в цій пропонується інформаційна технологія візуального аналізу рентгенівських зображень для пояснення результатів діагностування пневмонії. В основі технології закладено модель класифікації на основі згорткової нейронної мережі для вилучення слабо виражених ознак ранньої вірусної пневмонії та модифікований метод відмінної локалізації для інтерпретації результатів класифікації. Метод інтерпретації полягає в застосуванні зважених градієнтів до мап активації класів. Подібна модифікація забезпечує відмінну локалізацію аномальних зон на рентгенограмах, що дає змогу вилучити цільові слабко виражені ознаки ранньої пневмонії. Відповідно до обчислювальних експериментів, запропонована інформаційна технологія може бути ефективним засобом для миттєвого діагностування в разі перших підозр на виявлення пневмонії.
Информационная технология ранней диагностики пневмонии с использованием сверточных нейронных сетей
(2021) Радюк, Павло Михайлович; Радюк, П. М.; Radiuk, Pavlo M.; Бармак, О. В.; Крак, Ю. В.
За последние несколько лет пневмония стала одной из самых распространенных легочных. заболеваний во всем мире, а ее лечение является важнейшей задачей в клинической практике. Медицинский опыт доказал, что ранняя диагностика Пневмония является решающим фактором ее успешного лечения. На сегодняшний день автоматизированный рентгенографический анализ грудной клетки признан самым эффективным подходом в диагностике легочных заболеваний, в частности пневмонии. Однако до сих пор не ясно, какие пневмонические признаки на рентгеновском изображении автоматизированный метод диагностики относят к ранней стадии заболевания. Кроме того, вопрос интерпретирования результатов цифровой диагностики также не решен и требует дальнейшего изучения. Поэтому в представленной работе предлагается информационная технология визуального анализа рентгеновских изображений для интерпретирования результатов цифровой диагностики вирусной пневмонии на ранних стадиях. Технология включает модель классификации, на основе сверточной нейронной сети, для извлечения нечетких признаков ранней вирусной пневмонии и модифицированный метод отличной локализации для объяснения результатов классификации. Нейронная сеть, используемая в исследовании, содержит эффективную расширенную операцию свертки для объединения признаков из разных рецептивных полей на картинке. Предлагаемый метод интерпретации заключается в применении взвешенных градиентов к картам активации классов. По результатам вычислений ис- ристанная модель превзошла другие нейронные архитектуры по показателю precision (98,5%), но уступила accuracy (96,1%) и recall (93,6%). Кроме того, модель продемонстрировала сравнительно низкие значения ошибок первого и второго рода, достигнув 1,4 и 6,4% соответственно. В общем, согласно вычислительным экспериментов, предложенная информационная технология может быть эффективным инструментом мгновенной диагностики при первом подозрении на пневмонию.

Переглянути

Нові надходження