International Workshop on
Advances & Challenges in Computing (A2C)
A Regular meeting
Date: 18 March 2026 (Wednesday)
Time: 18:00-20:00 (East European time, Kyiv)
Format: Online
Place: Link to event
Email: a2c.workshop@gmail.com
AGENDA:
- Speakers: Roman Kvyetnyy, Oleh Bisikalo, Yevhen Palamarchuk, Olena Kovalenko
- Speakers: Roman Kvyetnyy, Yaroslav Ivanchuk, Oleksii Kozlovskyi
- Information about research projects and conferences: Anatoliy Sachenko
- Speakers: Roman Kvyetnyy, DSc in Engineering, Professor, Corresponding Member of NAES of Ukraine, Professor of the Department of Automation and Intellectual Information Technologies
Oleh Bisikalo, DSc in Engineering, Professor, Head of the Department of Automation and Intellectual Information Technologies
Yevhen Palamarchuk, PhD in Engineering, Associate Professor, Professor of the Department of Automation and Intellectual Information Technologies
Olena Kovalenko, PhD in Engineering, Associate Professor, Associate Professor of the Department of Software
Vinnytsia National Technical University, Vinnytsia, Ukraine
Main Speaker: Olena Kovalenko
E-mail: ok@vntu.edu.ua
Topic: Methodology for developing an educational electronic information environment.
Presentation in Ukrainian
Abstract: The methodology for developing an educational electronic information environment involves organizing the process into distinct stages, methods, models, and algorithms for constructing an informational Internet space for learning. It incorporates techniques for information reflection based on the mirror concept, the formation of electronic information environment contours, the interaction of space agents, and the establishment of a visual collaboration library tailored for adaptation to specific educational institutions or corporate training contexts.
This methodology is grounded in the principles of systems theory, set theory, graph theory, information systems, self-learning organizations, and visual collaboration. Mathematical and visual models facilitate the organization of data sets into unified formulas of association and interconnection, thereby presenting a structural information system that underpins the creation of an educational electronic information space and supports agent interaction. This environment serves as the foundation for an electronic university within an educational institution. In contrast to traditional platforms such as Moodle and Canvas, which primarily digitalize the learning process, the JetIQ+ system emphasizes the digitization and automation of educational, management, and support processes. JetIQ, as an information ecosystem, enables the implementation of specific educational and management processes and supports both methodological and scientific activities.
Specialized services facilitate various forms of communication, including student-teacher, student-student, student-group, group-teacher, teacher-dean’s office, and manager-subordinate (group of employees) interactions, as well as communications with software agents such as personal repositories, navigators, VNTU repositories, educational program management modules, and departmental Jet-sites. For conferences, cloud technologies are utilized, and integration with corporate email, Microsoft Office 365, and Google Workspace for Education tools is provided.
Scientific novelty of research:
A novel methodology for developing an educational electronic information environment is introduced. This methodology encompasses concepts and methods, models of agent interaction within the educational information ecosystem, and educational, managerial, and supporting processes. The approach enhances the quality of the design and implementation of educational electronic information environments and reduces the time required to adapt implementation methods to the specific requirements of individual educational institutions and corporate training programs.
A project management technology for implementing an educational electronic information environment is proposed. This technology enables the identification of requirements for individual modules, verification of agent interactions, assessment of error risks in specific processes, and evaluation of the socio-technical level of the educational information ecosystem. These features collectively reduce the time required to implement new modules and changes within the educational electronic information environment.
A visual collaboration library is proposed that, unlike existing solutions, employs specialized visual elements to reduce facilitation time and to define requirements for developing an educational electronic information environment. This library also enables rapid visual modeling of educational, management, and support processes.
The TestIQ testing system has been enhanced to incorporate artificial intelligence technologies, distinguishing it from existing systems. As a result, the time required to prepare test tasks is reduced, test results are automatically evaluated, coverage of theoretical and practical educational content is improved, and teacher workload is reduced.
The theory of information reflection for educational electronic information environments has been refined by clarifying post-processing information reflection processes, considering the characteristics of educational, managerial, and auxiliary processes. This refinement enables assessment of the socio-technical level of the information ecosystem at the strategic level and formalizes information flows within individual modules and agent interactions at the contour level. Consequently, input and output information can be structured in accordance with specified requirements and international standards for educational software, facilitating the optimization of the learning space based on defined criteria.
The structural model of the educational electronic information environment has been enhanced to incorporate functional, communication, and motivational-emotional contours within a unified system, distinguishing it from existing models. This approach enables the separation of specific functions, reduces the prominence of the functional contour, and optimizes communication and motivational-emotional processes in accordance with specified criteria.
The model for creating an educational electronic information environment for clients has been improved, in contrast to educational spaces in CRM systems. This adaptation of the contour model of training is designed to enhance client loyalty.
The model for developing an educational electronic information environment for microlearning has been enhanced using the modules “Electronic book” and “Navigator of academic disciplines.” Unlike existing models, it incorporates SMART elements that improve communication and motivational-emotional contours by increasing the number of short motivating messages for independent student work, utilizing test tasks, and automating the evaluation of student work results, thereby reducing teacher workload.
The general methodology is applicable to any educational institution and can be further adapted to meet specific needs. Additionally, methods for developing open educational courses are proposed, including microlearning and mobile application development. The methodology facilitates the identification of correspondences among pedagogical theories, methodologies, and technologies to create an educational electronic information environment based on a multi-agent information ecosystem. This approach enables the parallel adaptation of technologies during implementation.
Resume: Candidate of Technical Sciences, Associate Professor of the Department of Software, Faculty of Information Technologies and Computer Engineering.
Specialist of the Center for Digitalization of the Educational Process of VNTU,
Doctoral Candidate of the Department of AIIT of VNTU.
Vinnytsia National Technical University (VNTU).
Education – VNTU, 1983, specialty “Electronic Computing Machines”, Candidate of Technical Sciences since 1997, dissertation topic.
“Development and Research of Self-Calibrating Computing ADCs and DACs for the System of Digital Processing of Analog Information”. Since 2000, he has been researching and implementing distance and blended learning at Vinnytsia National Technical University; Vinnytsia National Agrarian University, Vinnytsia Trade and Economic Institute of Kyiv National Trade and Economic Institute.
Co-author of methodological developments on the development of an information learning environment at international exhibitions (gold medal of the All-Ukrainian Exhibition “Innovation”, gold medal of the exhibition “Education and Career”); diploma for presentations at exhibition events; diploma for the development of innovative technologies in education of the Ministry of Agrarian Policy, the Ministry of Education and Science of Ukraine.
Coordinator of the Ecommis+ project (VNAU) (Ukraine, Lithuania, Israel, Germany).
Author and teacher of trainings on the topics of “Project Planning”; “Scrum Methodology”, “SMART Planning”, etc.
Co-coordinator and teacher of advanced training courses “Creating electronic resources for blended learning of students in the environment of JetIQ educational process support systems”.
Co-author of the project for creating an educational environment in Vinnytsia region (2019).
Developer of the JetIQ project of VNTU. “Electronic University” since 2015 for the present.
Has over 320 publications
GOOGLE Scholar H-index 12
ORCID 0000-0003-2864-9058
Scopus https://www.scopus.com/authid/detail.uri?authorId=57198156271
H-index 4
- Speakers: Roman Kvyetnyy, DSc in Engineering, Professor, Corresponding Member of NAES of Ukraine, Professor of the Department of Automation and Intellectual Information Technologies
Yaroslav Ivanchuk, DSc in Engineering, Professor, Professor of the Department of Computer Science
Oleksii Kozlovskyi, Postgraduate student, Department of Computer Science
Vinnytsia National Technical University, Vinnytsia, Ukraine
Main Speaker: Yaroslav Ivanchuk
E-mail: ivanchuck@ukr.net
Topic: Structural Optimisation of Regression Models Based on a Genetic Algorithms.
Presentation in Ukrainian
Abstract: The article presents the results of research into methods of structural optimisation of regression models based on a genetic algorithm. The main focus is on solving the problem of automated selection of informative features and optimisation of the structure of regression models in order to increase the accuracy of forecasting and improve their generalisation ability. The proposed approach considers structural optimisation as a global search problem in a multimodal space of possible model structures. This determines the expediency of using evolutionary optimisation methods.
The key parameters of the genetic algorithm are set at the preliminary stage using a reference nonlinear test function, which is characterised by the presence of a known structure of relevant features and complex nonlinear interactions between them. This allows for an objective assessment of the genetic algorithm’s ability to perform global optimisation and ensures the correct selection of its operating parameters for further application in nonlinear regression problems.
The effectiveness of the proposed approach was investigated using a number of common regression analysis methods, including Gradient Boosting Regressor, Random Forest, k-nearest neighbours, linear regression, and a neural network based on a multilayer perceptron. Experimental studies were conducted using the California Housing reference dataset. The results of the experiments indicate an increase in the accuracy of predictions for the Gradient Boosting Regressor, Random Forest, and k-nearest neighbours models due to the application of structural optimisation based on a genetic algorithm. At the same time, linear regression and neural networks demonstrated high robustness to changes in the composition of input features, which is due to the peculiarities of their internal structure and learning mechanisms.
In general, the results obtained confirm the feasibility of using genetic algorithms as an effective tool for structural optimisation of regression models in machine learning and intelligent data analysis tasks, as well as their potential for building models with increased accuracy and generalisation ability.
Scientific novelty of research:
A method for structural optimisation of regression models has been developed which, unlike existing approaches, is based on the use of evolutionary computation involving static and evolutionary penalty functions. The proposed method provides targeted control of model structure complexity in the evolutionary search process by adaptively balancing approximation accuracy and structural compactness. The use of static penalties allows a priori constraints on the model structure to be fixed, while evolutionary penalty functions change dynamically according to the state of the population and the stage of optimisation, which increases the efficiency of the global search and the stability of the method against premature convergence.
An approach has been developed for identifying the working parameters of the structural optimisation method for regression models based on a genetic algorithm, which, unlike existing approaches, is based on the use of a reference nonlinear function that forms a multimodal search space with a predetermined structure of relevant features. This provides the possibility of objectively assessing the ability of evolutionary methods for global optimisation and automated detection of the optimal structure of a regression model.
Resume: DSc in Engineering, Professor, Professor of the Department of Computer Science, Faculty of Intelligent Information Technology and Automation.
Vinnytsia National Technical University (VNTU).
Education – VNTU, 2005, specialty «Automobiles and the automotive industry», Doctor of Technical Sciences since 2021, dissertation topic: «Methods and means of mathematical modelling of hydraulic vibration and vibration-impact machines». Since 2020, he has been an active member of the organising committee of the Sikorsky Challenge Start-up School. Research area: «Mathematical modelling of technical systems and intelligent data analysis». Member of the project team responsible for training higher education students in the field of Computer Science. Since 2024, he has been the academic secretary and member of the specialised academic council D 05.052.01 for the defence of doctoral dissertations at Vinnytsia National Technical University. Member of the editorial board of the International Scientific and Technical Journal «Information Technologies and Computer Engineering».
The result of scientific and pedagogical activity is more than 250 publications, including: 6 monographs, 2 of which were published abroad by Taylor & Francis Group, CRC Press, Balkema book (London, United Kingdom), 5 teaching aids, 3 textbooks, 22 Ukrainian patents, 80 articles in professional publications, including 13 articles in publications included in the Scopus and Web of Science scientometric databases.
More than 250 publications
GOOGLE Scholar (H-index 15)
ORCID: 0000-0002-4775-6505
Scopus Author ID: 57170734800 (H-index 5)
Web of Science ResearcherID: J-2797-2018
- Information about research projects and conferences: Anatoliy Sachenko
Doctor of Technical Sciences, Professor, scientific advisor of Research Institute for Intelligent Computer Systems.
E-mail: as@wunu.edu.ua