To obtain access to full text of journal and articles you must register!
- Article name
- Development of anti-fraud tools in banking systems
- Authors
- Ishchanova S. G., , ischanovaSG@yandex.ru, Institute of Physics and Technology (National Research University); SC "OKB SAPR", Moscow Region, Dolgoprudny, Russia; Moscow, Russia
- Keywords
- anti-fraud / banking systems / KAN / dataset collection / data synthesis / behavioral data
- Year
- 2026 Issue 2 Pages 76 - 86
- Code EDN
- EIQFSI
- Code DOI
- 10.52190/2073-2600_2026_2_76
- Abstract
- Modern banking systems actively use anti-fraud tools, but at the moment their effectiveness is insufficient. A contradiction has arisen related to the need to prevent financial losses through the use of anti-fraud tools and the lack of scientifically based methods for identifying fraudulent transactions. To resolve the contradiction, this work proposes to apply neural network methods for evaluating operations, expanding the feature space of the dataset, which will bring the detection of fraudulent operations closer to real time, as well as reduce the number of erroneous blockings of operations. During the work, a dataset containing data on user behavior when working with the application was collected, a synthesis of data based on real data was carried out, and a comparative analysis of neural network models was presented from the point of view of solving the antifraud problem. It is shown that, based on an extended data set, it is possible to solve the antifraud problem close to real time. Further directions of research are identified, in particular, improving the quality of synthesis models, and the connection of the work with research in behavioral economics, which models the behavior of subjects of economic relations using large language models, is noted.
- Text
- BUY for read the full text of article
- Buy
- 500.00 rub
