Machine Learning for Fraud Detection in SaaS Platforms
DOI:
https://doi.org/10.36676/urr.v12.i1.1466Keywords:
Machine learning, fraud detection, SaaS platforms, deep learning, anomaly detection, ensemble models, feature engineering, transfer learning, federated learning, explainable AI, privacy-preservingAbstract
Fraud detection in Software as a Service (SaaS) platforms has garnered significant attention in light of the growing complexity of cybercrimes and the growing need to protect sensitive user data. Although traditional methods of fraud detection are largely based on rule-based systems, machine learning (ML) has emerged as a more effective option due to its ability to detect complex and dynamic patterns of fraud. This paper performs a literature review for the period 2015-2024, examining the use of various ML techniques in fraud detection in SaaS environments. Early research focused on basic classifiers like decision trees and logistic regression, gradually moving towards ensemble methods and feature engineering to achieve higher accuracy.
References
• Khatri, S., Agarwal, A., & Sharma, S. (2015). A study of fraud detection models in e-commerce and SaaS environments using machine learning techniques. Journal of Data Science and Analytics, 8(3), 125-142.
• Zhang, Y., & Wang, Z. (2016). Ensemble learning techniques for fraud detection in SaaS platforms. International Journal of Computer Applications, 45(6), 12-19.
• Jang, H., Lee, T., & Park, Y. (2017). Unsupervised learning for anomaly detection in cloud-based services. Cloud Computing and Applications Journal, 22(1), 45-56.
• Gupta, R., & Soni, N. (2017). Support vector machine based fraud detection in SaaS platforms: A case study in the subscription model. International Journal of Machine Learning and Computing, 6(4), 295-305.
• Chen, L., Zhang, H., & Liu, Z. (2018). Feature engineering and selection in machine learning models for fraud detection in cloud-based services. Data Science Review, 3(2), 98-107.
• Li, J., & Wang, X. (2019). Hybrid machine learning models for fraud detection in SaaS applications: A comparison of ensemble methods and neural networks. Journal of Artificial Intelligence and Applications, 11(5), 52-63.
• Rehman, M., Khan, F., & Rizwan, M. (2020). Using isolation forests for fraud detection in SaaS platforms with high-dimensional data. Journal of Cloud Security and Fraud Prevention, 5(3), 200-211.
• Singla, A., Sharma, D., & Singh, P. (2020). Deep learning for real-time fraud detection in cloud-based platforms. Neural Networks and Machine Learning Journal, 32(4), 411-426.
• Yang, C., & Xu, B. (2021). Reinforcement learning-based fraud detection in SaaS environments: A dynamic approach. Journal of Computational Intelligence in Security, 15(6), 94-106.
• Zhao, F., & Zhang, S. (2021). Q-learning for fraud prevention in multi-tenant SaaS platforms. International Journal of Cloud Computing and Services Science, 9(2), 37-49.
• Lee, Y., Kim, H., & Lee, J. (2022). Federated learning for privacy-preserving fraud detection in decentralized SaaS platforms. Journal of Privacy and Security Technology, 13(1), 54-67.
• Sharma, A., & Singh, R. (2023). Federated learning for scalable fraud detection in global SaaS environments: A comparative study. International Journal of Distributed Computing and Machine Learning, 8(5), 118-129.
• Kumar, S., & Rathi, A. (2024). Explainable AI for fraud detection: Making machine learning models transparent and interpretable for SaaS applications. Journal of AI Ethics and Transparency, 17(2), 207-220.
• Singh, R., & Verma, A. (2024). Hybrid machine learning models for fraud detection: Enhancing accuracy with interpretability for SaaS platforms. Journal of Machine Learning and Computing, 12(3), 56-72.
• Jang, D., & Li, Q. (2024). Multi-modal fraud detection systems: Integrating transactional, behavioral, and device data in SaaS platforms. Journal of Cloud-Based Fraud Prevention and Security, 10(1), 89-103.
Downloads
Published
How to Cite
Issue
Section
License
Copyright (c) 2025 Universal Research Reports

This work is licensed under a Creative Commons Attribution-NonCommercial 4.0 International License.