Registration Number
0225U005122,
(0124U002504)
, R & D reports
Title
To develop generalized optimization methods for intelligent data analysis and their infrastructures
popup.stage_title
Розробити узагальнені оптимізаційні процедури прийняття рішень на частково визначених даних та інфраструктурах
Head
Stovba Viktor O., Доктор філософії
Registration Date
25-12-2025
Organization
V.M. Glushkov Institute of Cybernetics of the National Academy of Sciences of Ukraine
popup.description1
The goal of scientific research is to develop of generalized optimization methods on partially ordered sets and structured data
popup.description2
For finding parameters of linear regression model with L1-regularization and a criterion based on the sum of absolute deviations of predicted values from the observed ones powered to p={1,2}, the emlmpr algorithm is developed. For finding unknown parameters of a generalized nonsmooth regression model, the empq algorithm is proposed, which enables noise removal from data and reconstruction of the original signal. Models and methods of group decision-making and their application to generalized group cybernetic systems are investigated. The challenges that arise when extending the analysis of individual decisions to group decision-making are examined. Mixed strategies of cooperation and generalized leadership are presented, along with the main theorems characterizing the behavior of participants under these conditions and the desired performance indicators, in particular their output and profit. Necessary and sufficient conditions for consistency of constraint systems are established and justified for two types of transportation problems: transportation problem with two-sided constraints on unknown consumer demands, and two-stage transportation problem with two-sided constraints on unknown consumer demands and intermediate points capacities. These conditions allows one to verify feasibility prior to solving the problem (in particular, using solvers available on NEOS server) and thus to save computational and time resources when dealing with large-scale data. Data exchange in supply chain networks is analyzed, and evolution of the data exchange concept toward a new concept of data ecosystems is outlined. It was shown that data spaces represent one of the most promising approaches to implementation of successful industrial data ecosystems. A mathematical model based on a linear regression model and the accelerated subgradient algorithm are developed for anomaly detection in cattle behavioral data, enabling identification of both typical and anomalous behavioral patterns.
Product Description
popup.authors
Maksym S. Dunaievskyi
Seit-Bekir S. Suleimanov
popup.nrat_date
2025-12-25
