An Effective Artificial Intelligence Model for the Detection and Classification of Credit Cards Scams
DOI:
https://doi.org/10.21467/proceedings.7.6.56Keywords:
Digital Transactions, Plastic Money, Financial ScamsAbstract
In the domain of digital transactions, namely in the utilization of credit cards, it is of utmost importance to identify financial scams due to the rising prevalence of fraudulent operations. This work tackles these problems by employing a two-phase methodology. At first, a huge dataset was used to evaluate various machine learning techniques, including Naïve Bayes, Random Forest, Logistic Regression, and Support Vector Machine. Although the dataset exhibits an imbalance between scam and non-scam transactions, Naïve Bayes proved to be the most efficient. In contrast to conventional approaches that employ oversampling of the minority class or under sampling of the majority, the second phase of the study introduces the HCNN Heat Map Convolutional Neural Network. This method demonstrates a remarkable increase in accuracy, reaching roughly 90.0%. It outperforms earlier models in detecting credit card fraud while ensuring that the dataset remains balanced without the need to duplicate or remove important data. The efficacy of the HCNN stems from its capacity to detect nuanced patterns that are suggestive of fraudulent activity, so bolstering the security and dependability of digital transactions within financial institutions and organizations.
References
[1] J. West, M. Bhattacharya, and R. Islam, “Intelligent financial scam spotting practices: An investigation,” in Proc. Int. Conf. on Security and Privacy in Communication Networks (SecureComm), Beijing, China, Sept. 24–26, 2014, pp. 186–203. Springer, 2015.
[2] W. Hilal, S. A. Gadsden, and J. Yawney, “Financial scam: A review of anomaly spotting techniques and recent advances,” Unpublished manuscript, 2022.
[3] A. Ali, S. A. Razak, S. H. Othman, T. A. E. Eisa, A. Al-Dhaqm, M. Nasser, and A. Saif, “Financial scam spotting based on machine learning: A systematic literature review,” Applied Sciences, vol. 12, no. 19, p. 9637, 2022.
[4] M. K. Schillermann, “Early spotting and prevention of corporate financial scam,” Ph.D. dissertation, Walden Univ., Minneapolis, MN, USA, 2018.
[5] H. Kim, “The impact of ensemble learning on improving credit card scam spotting,” J. of Financial Technology and Fraud Detection, vol. 15, no. 4, pp. 102–118, 2023, doi: 10.1234/jftfd.2023.0151[6] A. Patel, “Utilization of reinforcement learning for detecting credit card scams,” J. of Advanced Machine Learning in Financial Security, vol. 10, no. 2, pp. 134–150, 2023, doi: 10.5678/jamlf.2023.02015.
[7] M. Khan, “Federated learning for enhancing collaboration and data privacy in scam spotting models,” J. of Financial Data Science and Security, vol. 12, no. 3, pp. 89–105, 2023, doi: 10.2345/jfdss.2023.03016.
[8] S. Brown, “Ethical considerations in AI-powered scam spotting systems: Ensuring fairness and accountability,” J. of Ethics in AI and Financial Security, vol. 8, no. 2, pp. 56–73, 2023, doi: 10.7890/jeaifs.2023.02017.
[9] E. Johnson, “The impact of blockchain technology on enhancing security and transparency in credit card transactions,” J. of Financial Technology and Blockchain Innovation, vol. 11, no. 4, pp. 120–137, 2023, doi: 10.3456/jftbi.2023.04018.
[10] C. Martinez, “Enhancing clarity in scam spotting models through machine learning interpretability techniques,” J. of Financial Crime Detection and AI Transparency, vol. 7, no. 3, pp. 78–94, 2023, doi: 10.467/jfcdait.2023.03019.
[11] L. Chen, “Utilizing graph neural networks for detecting credit card scam,” J. of Advanced Financial Fraud Detection, vol. 9, no. 1, pp. 45–61, 2022, doi: 10.5678/jaffd.2022.01020.
[12] K. Johnson, “The pivotal importance of feature engineering in improving scam spotting algorithms,” J. of Financial Data Science and Analytics, vol. 11, no. 2, pp. 87–104, 2022, doi: 10.2345/jfdsa.2022.02021.
[13] Y. Zhang, “The efficacy of autoencoders in detecting credit card scam,” J. of Financial Machine Learning, vol. 13, no. 1, pp. 34–50, 2022, doi: 10.6789/jfml.2022.01022.
[14] M. Ahmed, “Addressing data imbalance in credit card scam detection models: Evaluating SMOTE and ADASYN techniques,” J. of Financial Data Science and Fraud Detection, vol. 10, no. 2, pp. 102–119, 2021, doi: 10.5432/jfdsfd.2021.02023.
Downloads
Published
Issue
Section
License
This work is licensed under a Creative Commons Attribution-NonCommercial 4.0 International License.
