Hybrid Real-Time Fraud Detection in Finance

Ullasa Kumar, Jennifer (2025) Hybrid Real-Time Fraud Detection in Finance. Masters thesis, Dublin, National College of Ireland.

Preview	PDF (Master of Science) Download (1MB) | Preview
Preview	PDF (Configuration Manual) Download (1MB) | Preview

Abstract

Detection of fraudulent behavior in financial transactions requires accurate, real-time analytics as the number of payments grows and more refined fraudulent approaches evolve. Traditional batch-based processing systems suffer from limitations related to concept drift and slow feedback mechanisms. This paper presents a streaming architecture for credit card fraud identification, using Apache Kafka to enable high-throughput data consumption and Adaptive Random Forest (ARF) model for continuous online learning, reinforced with logistic regression as static comparative baseline. The system performs real-time transaction processing, feature extraction, and issuing of fraud alerts with low latency. The ARF model adapts to changing patterns using internal drift detection mechanisms(1), persistently outperforming the baseline in detection accuracy, drift resistance, and scalability during periods of high event rates. The Kafka-driven pipeline ensures low latency working with thousands of events per second. I perform statistical validation of the outcome and examine ethical considerations like privacy and fairness(2). This framework presents a scalable and efficient solution for detection of nonstationary fraud, with future research directions involving optimized drift detection methods, deep learning algorithms, and cost-sensitive learning approaches.

Item Type:	Thesis (Masters)
Supervisors:	Name Email Gyamfi, Eric UNSPECIFIED
Subjects:	H Social Sciences > HG Finance H Social Sciences > HG Finance > Credit. Debt. Loans. Q Science > Q Science (General) > Self-organizing systems. Conscious automata > Machine learning
Divisions:	School of Computing > Master of Science in Data Analytics
Depositing User:	Ciara O'Brien
Date Deposited:	27 Aug 2026 09:11
Last Modified:	27 Aug 2026 09:11
URI:	https://norma.ncirl.ie/id/eprint/9679

Actions (login required)

View Item