Synergy of quantum computing and federated learning in information technology detection of hidden target groups

• Date: 2025
• Main Authors: Fadeichev, Serhii, Trofymchuk, Oleksandr, Trysnyuk, Vasyl
• Format: Article
• Language: Ukrainian
• Published: Kyiv National University of Construction and Architecture 2025
• Subjects: information technologies; federated learning; quantum computing; hybrid algorithms; big data analysis; hidden target groups
• Online Access: https://es-journal.in.ua/article/view/351681
• Journal Title: Environmental safety and natural resources

The article considers the information technology of detecting hidden target groups in large distributed data sets based on the synergy of quantum computing and federated learning. It is shown that in modern information systems for analyzing big data, traditional methods of machine learning and intelligent analysis demonstrate limited effectiveness in identifying poorly represented or hidden structures, especially under conditions of statistical imbalance of samples, high dimensionality of the feature space and distributed storage of information. The proposed approach combines federated learning as an information technology of distributed model formation without data centralization with quantum algorithms of amplitude amplification, which allows to increase sensitivity to weak signals in subspaces with increased information significance. The paper considers the architecture of a hybrid information technology, which includes classical computing nodes, a level of federated aggregation and a quantum computing module, as well as a structural and algorithmic scheme of interaction of classical and quantum components. The algorithmic aspects of the implementation of the proposed approach are analyzed, taking into account the limitations of modern quantum computing platforms of the NISQ (Noisy Intermediate-Scale Quantum) class, in particular the limited number of qubits and the influence of noise. It is shown that the use of selective quantum processing and iterative interaction between classical and quantum circuits allows to ensure the reproducibility of the results and the possibility of practical implementation of the proposed information technology. The proposed approach is a promising direction in the development of information technologies for analyzing large distributed data and can be used as a theoretical and applied basis for further research in the field of hybrid classical-quantum computing systems.