Credit Card Fraud Detection: A Hybrid Approach

Kolawole, Damilare Abel (2023) Credit Card Fraud Detection: A Hybrid Approach. Masters thesis, Dublin, National College of Ireland.

Preview	PDF (Master of Science) Download (2MB) | Preview
Preview	PDF (Configuration Manual) Download (4MB) | Preview

Abstract

The advancement of technology and the widespread use of online transactions have had a tremendous influence on the financial system, resulting in an increase in credit card-related fraud. This research looks at the effectiveness of a Hybrid Deep Learning Approach, especially an Autoencoder-Long Short-Term Memory (LSTM) model, in dealing with the problem of unbalanced datasets in credit card transactions. The study delves into two critical questions: first, how to effectively train a deep learning model on imbalanced datasets where legitimate transactions far outnumber fraudulent ones, thereby benefiting financial institutions, businesses, and cardholders; Second, it compares the proposed Hybrid Deep Learning Approach to current models in credit card fraud detection, with the goal of improving detection systems for different stakeholders. The research focuses on the unbalanced nature of credit card transaction datasets by using the Synthetic Minority Over-sampling Technique (SMOTE) for dataset balancing and feature selection. The hybrid deep Learning Approach incorporates an autoencoder to compress and extract key features, followed by an LSTM model to capture temporal relationships and sequential patterns in the data. This method improves anomaly detection by successfully discriminating irregular sequences. The results show that the hybrid model outperformed current methods in credit card fraud detection. The use of autoencoder-LSTM architecture allows the model to recognize abnormalities with greater precision and accuracy. Furthermore, visual representations such as ROC curves and confusion matrices demonstrate the model resilience, with higher Area Under the Curve (AUC) ratings.

Item Type:	Thesis (Masters)
Supervisors:	Name Email Tomer, Vikas UNSPECIFIED
Uncontrolled Keywords:	Credit Card Fraud; Machine Learning; Deep Learning; Class Imbalance; Detection; SMOTE
Subjects:	H Social Sciences > HV Social pathology. Social and public welfare > Criminology > Crimes and Offences Q Science > QA Mathematics > Electronic computers. Computer science T Technology > T Technology (General) > Information Technology > Electronic computers. Computer science H Social Sciences > HG Finance > Credit. Debt. Loans. H Social Sciences > HF Commerce > Electronic Commerce 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:	14 May 2025 14:00
Last Modified:	14 May 2025 14:00
URI:	https://norma.ncirl.ie/id/eprint/7550

Actions (login required)

View Item