Registration Number
0225U000371,
(0123U103045)
, R & D reports
Title
Developing of models of increasing usefulness and self-reinforcing for decentralized big data analysis systems
popup.stage_title
Розробити нові моделі аналізу великих цифрових даних у децентралізованих системах
Head
Kuznietsov Vladyslav O., к.т.н.
Registration Date
10-01-2025
Organization
V.M. Glushkov Institute of Cybernetics of the National Academy of Sciences of Ukraine
popup.description1
The goal is to develop models of self-reinforcing effects, their study, in the context of manifestations of synergistic and network effects in big data processing systems, as well as research (analysis) of large decentralized big data processing systems. As part of the research, it is envisaged to study the influence of the actions of system users and the amount of information that forms big data and affect the general informativeness of the data stored in the big data system in the format of metadata, documents, facts and other indicators. The work will explore models of self-reinforcing effects and increasing usefulness in big data within the process of data extraction, processing and use; new models of self-reinforcing and utility-enhancing effects and a model of big digital data analysis will be built.
The main tasks of the work are as follows: it is necessary to develop machine learning models on big data to identify network and synergistic effects that affect the value of data and, accordingly, demand through self-reinforcing effects. There is also a need to develop machine learning models that engage statistical metrics that are informative in the context of data value and demand growth. It is necessary to develop approaches for obtaining features from data, taking into account the hidden properties of data and their spatial location. Investigate the spatial location of features, assess anomalies and deviations in the data. Develop methods of visual data analysis and machine learning, taking into account user interactions and the big data platform. Develop big data models that take into account the criteria of value and usefulness of data obtained from various sources and entering the big data processing chain.
popup.description2
During the reporting period, the following results were obtained: 1) an approach to modeling and analyzing big data systems with independent behavior of individual nodes was developed by modeling the spatio-temporal dynamics of demand and supply in a big data network; 2) methods for modeling big data systems with an open control loop were improved by using approaches, models, and methods that implement optimal data demand management; 3) A study of the proposed approaches and methods was conducted, and the use of methods for analyzing large networks and resource allocation, in particular, models for distributing radio frequencies and transmitted signal power, and constructing optimal routes; 4) A study was conducted and the effectiveness of the proposed approaches and methods for optimal control was compared with other methods, in particular, methods involving dynamic neural networks. The obtained models, approaches, and algorithms can be used to formalize the spatio-temporal dynamics of demand and supply in big data networks, big data analysis systems, in particular, e-commerce, resource distribution systems (goods or services), modeling the behavior of big data systems, optimizing architectures of big data models. The results obtained in this research work can be used in the big data industry for the development of decision support systems with the inclusion of a person (supplier or user) in the decision-making process, for visual analytics and decision-making systems for big data processing systems, including in user service systems and service quality assessment.
Product Description
popup.authors
Dunaievskyi Maksym S.
Kuznietsov Vladyslav O.
Suleimanov Seit-Bekir S.
popup.nrat_date
2025-01-10
