Continual Learning Systems for Adaptive Fraud-Pattern Detection in  High-Velocity Transaction Streams

Gopinathan Rathinavelu

Continual Learning Systems for Adaptive Fraud-Pattern Detection in High-Velocity Transaction Streams

Authors

Gopinathan Rathinavelu

Keywords:

Continual Learning, Fraud Detection, Concept Drift, Transaction Streams, Online Learning, Adaptive Machine Learning

Abstract

High-velocity financial transaction streams exhibit severe class imbalance, evolving fraudstrategies, delayed verification labels, temporal dependencies, and frequent changes inlegitimate customer behaviour. Conventional batch-trained fraud detection models can therefore experience progressive performance degradation

References

1. Andrés L. Suárez-Cetrulo, David Quintana, and Alejandro Cervantes, “A Survey on Machine Learning for Recurring Concept Drifting Data Streams,” Expert Systems with Applications, vol. 213, article 118934, 2023. [Publication issue is 2023; exclude this reference if “before 2023” means strictly ≤2022.] (ScienceDirect)

Downloads

Published

2024-04-29

How to Cite

Gopinathan Rathinavelu. (2024). Continual Learning Systems for Adaptive Fraud-Pattern Detection in High-Velocity Transaction Streams . Journal of Computational Analysis and Applications (JoCAAA), 33(4), 1239–1264. Retrieved from https://eudoxuspress.com/index.php/pub/article/view/5849