Real-time Financial Anomaly Detection in SAP ERP Systems Using Ensemble Learning

Authors: Surya Sai Ram Parimi

DOI: https://doi.org/https://doi.org/10.5281/zenodo.12805572

Short DOI: https://doi.org/gt442t

Country: India

Full-text Research PDF File:   View   |   Download

Abstract: Financial anomaly detection is paramount in SAP ERP systems to safeguard against fraud, errors, and operational inefficiencies. This survey paper explores the application of ensemble learning techniques for real-time anomaly detection within SAP ERP environments. Ensemble methods, such as Random For-ests, Gradient Boosting Machines, and Neural Network Ensembles, combine multiple models to enhance detection accuracy and adaptability to dynamic data streams. The paper reviews case studies illustrating the effectiveness of ensemble learning in detecting anomalies across diverse sectors, including fraud de-tection in procurement, revenue leakage prevention, and real-time transaction monitoring. Challenges specific to real-time anomaly detection in SAP ERP systems, such as data integration complexity and interpretability of model outputs, are discussed alongside proposed solutions and best practices. The conclusion highlights the pivotal role of ensemble learning in strengthening financial security, optimizing operational efficiency, and mitigating risks within SAP ERP systems. Future research directions focus on advancing anomaly detection algorithms and integrating AI-driven automation to enhance detection ca-pabilities further.

Keywords: Anomaly Detection, SAP Financial Operations, ERP System, Ensemble Learning

Paper Id: 230798

Published On: 2024-03-13

Published In: Volume 12, Issue 2, March-April 2024

Share this