Saved in:
| Main Author: | |
|---|---|
| Format: | Recurso digital |
| Language: | |
| Published: |
Zenodo
2026
|
| Online Access: | https://doi.org/10.5281/zenodo.20110248 |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
Table of Contents:
- <div class="box"> <div class="abs-text">The Real-Time Financial Anomaly Intelligence System is a machine learning-driven solution designed to detect and prevent fraudulent financial transactions. Financial fraud continues to cause significant monetary losses globally, as evolving fraud patterns consistently evade traditional rule-based detection systems. This paper presents a system that leverages exploratory data analysis and advanced feature engineering—including transaction velocity, spending behavior patterns, and geo-risk indicators—to uncover hidden anomalies in transaction data. Supervised machine learning models, specifically Logistic Regression, Random Forest, and XGBoost, are trained on the publicly available PaySim synthetic financial dataset to classify transactions as legitimate or fraudulent. The proposed system achieves an accuracy of 94%, a precision of 91%, a recall of 89%, and an F1-score of 90% on the imbalanced test set. By adapting to dynamic behavioral patterns, the system enhances real-time fraud detection and significantly reduces false positives. This intelligent anomaly detection framework provides a scalable and datadriven approach to strengthening financial security systems.</div> </div>