A Novel Framework for Financial Cybersecurity and Fraud Detection Using XAI-RNN-SGRU
Cyber threats involve unauthorized access, alteration, or deletion of private information, extortion, and disruption of business operations. Traditional network security methods need more scalability, data protection, and difficulty detecting advanced threats. The hybridization of Explainable Artifi...
Saved in:
| Main Author: | |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
IEEE
2025-01-01
|
| Series: | IEEE Access |
| Subjects: | |
| Online Access: | https://ieeexplore.ieee.org/document/11003905/ |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| _version_ | 1849310027455660032 |
|---|---|
| author | Smarajit Ghosh |
| author_facet | Smarajit Ghosh |
| author_sort | Smarajit Ghosh |
| collection | DOAJ |
| description | Cyber threats involve unauthorized access, alteration, or deletion of private information, extortion, and disruption of business operations. Traditional network security methods need more scalability, data protection, and difficulty detecting advanced threats. The hybridization of Explainable Artificial Intelligence (XAI) with Ridgelet Neural Network (RNN) and Soft Gated Recurrent Unit (SGRU) (XAI-RNN-SGRU) is introduced to address these challenges. This model enhances financial data prediction and cybersecurity through advanced preprocessing for handling missing values and data imbalance, and it also normalizes numerical features and encodes categorical variables. Enhanced Principal Component Analysis (EPCA) and Improved Fast Random Opposition-based learning Aphid Ant Optimization (AAO) are utilized for feature extraction and desired feature selection. Then, it combines the Ridgelet Neural Network with Soft Gated Recurrent Unit for accurate predictions and uses the Improved Homomorphic Encryption (IHE) process to reinforce data protection during computations. Comprehensive testing on three data sets reveals outstanding predictive performance with accuracies of 99.55%, 99.99%, and 99.95%, respectively. The introduced method improves data security, scalability, and efficiency in real financial applications. It outperforms current techniques with a bit error rate of 0.00004 and a data protection score of 98.03%, providing secure and reliable financial operations. In addition, it enhances efficiency in encryption and decryption, with rates of 9.25 and 8.30 seconds, respectively, and is exceptionally scalable for mass-market cybersecurity use. Such improvements offer a real-world solution for financial institutions, cybersecurity companies, and digital payment systems, guaranteeing strong fraud detection, improved data security, and enhanced operational efficiency compared to the current methodologies. |
| format | Article |
| id | doaj-art-306ede1bff994244a96333a86c7f9495 |
| institution | Kabale University |
| issn | 2169-3536 |
| language | English |
| publishDate | 2025-01-01 |
| publisher | IEEE |
| record_format | Article |
| series | IEEE Access |
| spelling | doaj-art-306ede1bff994244a96333a86c7f94952025-08-20T03:53:52ZengIEEEIEEE Access2169-35362025-01-0113881348815510.1109/ACCESS.2025.357021611003905A Novel Framework for Financial Cybersecurity and Fraud Detection Using XAI-RNN-SGRUSmarajit Ghosh0https://orcid.org/0000-0003-2943-5908Department of Electrical and Instrumentation Engineering, Thapar Institute of Engineering and Technology, Patiala, Punjab, IndiaCyber threats involve unauthorized access, alteration, or deletion of private information, extortion, and disruption of business operations. Traditional network security methods need more scalability, data protection, and difficulty detecting advanced threats. The hybridization of Explainable Artificial Intelligence (XAI) with Ridgelet Neural Network (RNN) and Soft Gated Recurrent Unit (SGRU) (XAI-RNN-SGRU) is introduced to address these challenges. This model enhances financial data prediction and cybersecurity through advanced preprocessing for handling missing values and data imbalance, and it also normalizes numerical features and encodes categorical variables. Enhanced Principal Component Analysis (EPCA) and Improved Fast Random Opposition-based learning Aphid Ant Optimization (AAO) are utilized for feature extraction and desired feature selection. Then, it combines the Ridgelet Neural Network with Soft Gated Recurrent Unit for accurate predictions and uses the Improved Homomorphic Encryption (IHE) process to reinforce data protection during computations. Comprehensive testing on three data sets reveals outstanding predictive performance with accuracies of 99.55%, 99.99%, and 99.95%, respectively. The introduced method improves data security, scalability, and efficiency in real financial applications. It outperforms current techniques with a bit error rate of 0.00004 and a data protection score of 98.03%, providing secure and reliable financial operations. In addition, it enhances efficiency in encryption and decryption, with rates of 9.25 and 8.30 seconds, respectively, and is exceptionally scalable for mass-market cybersecurity use. Such improvements offer a real-world solution for financial institutions, cybersecurity companies, and digital payment systems, guaranteeing strong fraud detection, improved data security, and enhanced operational efficiency compared to the current methodologies.https://ieeexplore.ieee.org/document/11003905/Cybersecurityfinancial transactionshomomorphic encryptiondata protectionscalability |
| spellingShingle | Smarajit Ghosh A Novel Framework for Financial Cybersecurity and Fraud Detection Using XAI-RNN-SGRU IEEE Access Cybersecurity financial transactions homomorphic encryption data protection scalability |
| title | A Novel Framework for Financial Cybersecurity and Fraud Detection Using XAI-RNN-SGRU |
| title_full | A Novel Framework for Financial Cybersecurity and Fraud Detection Using XAI-RNN-SGRU |
| title_fullStr | A Novel Framework for Financial Cybersecurity and Fraud Detection Using XAI-RNN-SGRU |
| title_full_unstemmed | A Novel Framework for Financial Cybersecurity and Fraud Detection Using XAI-RNN-SGRU |
| title_short | A Novel Framework for Financial Cybersecurity and Fraud Detection Using XAI-RNN-SGRU |
| title_sort | novel framework for financial cybersecurity and fraud detection using xai rnn sgru |
| topic | Cybersecurity financial transactions homomorphic encryption data protection scalability |
| url | https://ieeexplore.ieee.org/document/11003905/ |
| work_keys_str_mv | AT smarajitghosh anovelframeworkforfinancialcybersecurityandfrauddetectionusingxairnnsgru AT smarajitghosh novelframeworkforfinancialcybersecurityandfrauddetectionusingxairnnsgru |