Journal
Scientific and Technical Journal of Information Technologies, Mechanics and Optics
UDK004.89
Issue:3 (161)
Download PDF0 Kbyte
Detecting fraud activities in financial transactions using SMOTENN model
Annotation
The financial industry plays an important role in national economic growth. Because of their critical function, banks have become prime targets for numerous financial crimes. Among these, fraudulent financial transactions are regarded as a severe issue in the financial industry. Conventional approaches are frequently criticized for being ineffective in dealing with fraud in finance; therefore, machine learning approaches have a potential answer to deal with this problem. The goal of this research is to introduce a novel SMOTENN model to establish early detection of cyber fraudulent activities in financial transactions accurately. Two methods are used in this study: first, the Neural Network algorithm is applied to a dataset that contains unbalanced classes; second, the dataset is balanced using the SMOTE (Synthetic Minority Over-sampling Technique) algorithm first, followed by the Neural Network algorithm which we refer to as SMOTENN. The both models are assessed using evaluation metrics of Area Under the Curve, F1-score, precision, recall, specificity, accuracy, and processing time. The comparative analysis shows that the performance of the new SMOTENN model with a balanced dataset is significantly better than that of the neural network approach with an imbalanced dataset, implying that the new SMOTENN model is effective in detecting fraud activities in financial transactions.
Keywords
Постоянный URL
Articles in current issue
- Explainability and interpretability are important aspects in ensuring the security of decisions made by intelligent systems (review article)
- Intensification of sol-gel synthesis of Mn-containing MgO-Al2O3-ZrO2-SiO2 system materials
- Conformational properties of polymer brushes with aggrecan-like macromolecules under strong stretching conditions on a cubic lattice
- Atmospheric air-phase singlet oxygen generator for practical multifunctional applications
- Two-stage algorithm for underwater image recovery for marine exploration
- Analysis of the cryptographic strength of the SHA-256 hash function using the SAT approach
- Investigation of the possibility of using evolutionary algorithms for conditional generation of attributed graphs
- Analysis of the applicability of existing secret separation schemes in the post-quaternary era
- Deep learning-enhanced contour interpolation techniques for 3D carotid vessel wall segmentation
- A deep learning approach for adaptive electrocardiogram-based authentication in an internet of things enabled telehealth system
- Method for identifying the active module in biological graphs with multi-component vertex weights
- Modeling of nonlocal porous functionally graded nanobeams under moving loads
- Design of the microelectromechanical logic element based on a comb-drive resonator
- Critical loads of antisymmetric and mixed forms of buckling of a CCCC-nanoplate under biaxial compression
- On the properties of compromise M-estimators optimizing weighted L2-norm of the influence function
- Combined approach to fault detection in complex technical systems based on bond-graph model
- Feature extraction methods for metagenome de Bruijn graphs collections based on samples classification information
- Automatic calibration of the receiving line of information and control systems in real time
- Model for storing spatial data of tensor geophysical fields
- Boundary estimation of the reliability of cluster systems based on the decomposition of the Markov model with limited recovery of nodes with accumulated failures