Yurii Futryk

Yurii Futryk

Auto-generated profile, based on publicly available information

Yurii Futryk

Національний університет «Львівська політехніка

Research Focus & Metrics

Academic Career Length

5Years (2022-Present)

(Calculated once Scilit publication data is synced: current year − year of first publication + 1)

Representative Publications

View All Publications
15 June 2022 in Computer systems and network
Verified
Journal Article
Closed access
LULviv Polytechnic National University
Lviv Polytechnic National University
ASAndriy Senyk
Andriy Senyk
OMOleksandra Manziy
Oleksandra Manziy
YFYuriy Futryk
Yuriy Futryk
OSOleksandr Stepanyuk
Oleksandr Stepanyuk
IMInstitute for Applied Problems of Mechanics and Mathematics
Institute for Applied Problems of Mechanics and Mathematics
Due to large-scale changes in the economy in the world and in Ukraine in particular, there has recently been a significant increase in interest in the problems of investment theory. An example is the intensification in recent years of the purchase of shares of large international companies and cryptocurrencies and, according to the rapid growth of their values. It is known that as a special case, the theory of investment considers the task of optimizing investment portfolios. It is established that the issue of decision-making on the formation and optimization of the investment portfolio is in the field of attention of both large investment companies and private investors, because choosing among possible alternatives for allocating investments within the financial assets market, the investor will get different results. It is accepted that the optimal distribution of the investment portfolio should provide the best return while maintaining the least risk, and the result should be understood as the amount of income received during the period of ownership of the investment portfolio. An information system to support the decision-making of the securities portfolio has been developed, which allows potential investors to independently on assess the effectiveness of the investment portfolio by comparing the growth dynamics of shares available on the financial market. It is known that most of the information encountered by the investor is in tabular format, and according to the methodology of scientific knowledge, people are more receptive to visualized ways of presenting information. The newly created information system uses a visualization process that presents available tabulated information in a structured form of diagrams, graphs, charts.

30 August 2023
Verified
Conference Paper
Closed access
TDTaras Dubyniak
Taras Dubyniak
OMOleksandra Manziy
Oleksandra Manziy
TGTomasz Gancarczyk
Tomasz Gancarczyk
ASAndriy Senyk
Andriy Senyk
YFYurii Futryk
Yurii Futryk
The development technologies and functional capabilities of the information system created by the authors aimed at decision-making support of the formation of a set of securities, which enables potential investors with little experience to assess independently the effectiveness of the investment portfolio by simulating the dynamics growth of assets available on the financial market are described. The proposed information system uses visualization methods that present available tabular information in the structured form of schemes, graphs, and charts. The web-oriented solution provides an opportunity to analyze and forecast portfolios in real time based on the available types of shares of various companies.

20 February 2026 in Grail of Science
Verified
Journal Article
Open access
ЮФЮрій Футрик
Юрій Футрик
У тезах подано практичну методологію препроцесингу та формування ознакового простору для прогнозування часових рядів моделями LSTM (Long Short-Term Memory). Навчальні набори формуються у вигляді матриці , де L — довжина вікна, d — кількість ознак; усі перетворення виконуються причинно-коректно, без витоку інформації (data leakage) під час масштабування та відбору ознак.

20 April 2026 in Information Technology and Computer Engineering
Verified
Journal Article
Open access
YFYurii Futryk
Yurii Futryk
IPIvan Peleshchak
Ivan Peleshchak
Time series forecasting is crucial for supporting decisions in financial analytics, where data is characterised by non-linearity, non-stationarity, and high noise levels. The purpose of the study was to determine the effective configuration of a recurrent neural network with a parallel combination of Long Short-Term Memory (LSTM) cell stacks to improve the accuracy of stock price forecasting, and the possibilities of applying the back-end model in industry, energy, and related domains. The study applied deep learning methods using the TensorFlow/Keras library, and used historical data from Google shares to train the model. It was established that the architecture with parallel-stacked blocks provided higher learning stability compared to standard recurrent models due to more efficient allocation of technical features of the time sequence. It has been experimentally proven that the optimal number of neurons in the hidden layers for such a task was 100-200 units, while a further increase in the power of the model lead to a retraining effect. It was found that the use of dropout regularisation in the range of 0.1-0.2 minimised the error in the validation sample, while values over 0.3 significantly slowed down the convergence of the algorithm. Feature analysis showed that integrating an exponential moving average with a short time window improved the model result, showing a higher correlation with the target index than the relative strength index. The prediction quality of the model was evaluated by the Mean Squared error (MSE), the Root Mean Squared Error (RMSE), and the Mean Absolute Percentage Error (MAPE). It was found that configurations (50-100 blocks) were characterised by increased MAPE values, while in the range of 180-400 blocks the error decreased and became stable. The most accurate result was obtained for a configuration with 325 blocks, Dropout regularisation = 0.05 and Nadam optimiser (Nesterovaccelerated Adam): MAPE=1.62%, RMSE=2.41, MSE=6.05. The practical significance of the study lied in the formulation of clear recommendations for setting up hyperparameters of LSTM models for applied short-term forecasting of financial series

22 February 2022
Verified
Conference Paper
Closed access
ASAndriy Senyk
Andriy Senyk
VPViktor Pabyrivskyi
Viktor Pabyrivskyi
OUOksana Ukhanska
Oksana Ukhanska
YFYuriy Futryk
Yuriy Futryk
YSYuliya Senyk
Yuliya Senyk
OSOleksandr Stepanyuk
Oleksandr Stepanyuk
The methodology of choosing the parameters of strengthening the structural elements of the bearing surfaces of aircraft is presented in the article. Using mathematical methods of deformable solids, a mathematical model of processing elements of aircraft is built, which will be much more resistant to external mechanical impact. To give the surface water-repellent and anti-icing properties, surface treatment with highly concentrated energy flows can be used, followed by its strengthening. Such technologies include laser irradiation, plasma flux, electron beam, light beam, and ion flux. The proposed model is based on the problem of determining the technological parameters of the impact of concentrated energy flow on the surface of the part in order to determine with high accuracy the area of hardening and forecasting of zones of plastic deformation. The precision of determining these parameters is justified by taking into account the problem of dependence of physical and mechanical characteristics of the material on temperature - thermal sensitivity of the material. As these technologies involve significant changes in temperature values, taking into account temperature dependences is an important factor. The proposed methodology of determining the parameters is an actual task and is based on the solution of nonlinear problems, as well as allows you to create more accurate models of physical processes and provides reliability.

30 January 2025 in Herald of Khmelnytskyi National University. Technical sciences
Verified
Journal Article
Open access
PDF file
ІПІван Пелещак
Іван Пелещак
ЮФЮрій Футрик
Юрій Футрик
Удосконалення методів прогнозування часових рядів є важливим завданням для багатьох галузей, таких як фінанси, виробництво, військова справа, медицина та енергетика. Особливо актуальним є використання рекурентних нейронних мереж з блоками LSTM, які дозволяють ефективно враховувати довготривалі залежності у даних. Проте оптимальні архітектури LSTM та параметри, такі як кількість блоків і рівень Dropout, залишаються недостатньо дослідженими. Розробка та оптимізація морфології рекурентної нейронної мережі для прогнозування часових рядів з використанням LSTM блоків, інтеграція EMA (exponential moving average), RSI (relative strength index) та дослідження впливу різних параметрів моделі на точність прогнозування. У цій роботі застосовано метод оптимізатора Nadam для оптимізації морфології LSTM, який включає порівняння моделей з різними конфігураціями блоків LSTM (300–350 блоків), параметрів Dropout. Для навчання використовувались фінансові дані акцій Google (GOOGL), зібрані за допомогою бібліотеки yfinance, та метрики оцінювання точності, такі як MSE, RMSE, MAE, MAPE. Запропоновано оптимальну архітектуру LSTM для прогнозування часових рядів, яка враховує використання технічних індикаторів та оптимізатора Nadam. Дослідження показало, що модель з 350 блоками LSTM і Dropout 0.05 досягла мінімальної похибки 1.64% MAPE, що менша за результати попередніх досліджень. Запропоновані морфологічні та архітектурні рішення можуть бути використані для прогнозування цін акцій, обсягів продажів та задач в різних галузях, де потрібно прогнозування та аналіз часових рядів.