Cost – Sensitive Hyperparameter Tuning for XGBoost in Credit Card Fraud Detection

Nguyen, Thi Thuy Hang (2025) Cost – Sensitive Hyperparameter Tuning for XGBoost in Credit Card Fraud Detection. Masters thesis, Dublin, National College of Ireland.

Preview	PDF (Master of Science) Download (525kB) | Preview
Preview	PDF (Configuration Manual) Download (411kB) | Preview

Abstract

Credit card fraud detection is a daunting task considering the highly imbalanced nature of transaction data and the cost of false negatives being prohibitive. Traditional machine learning classifiers fail to perform well in such a scenario, leading to poor minority class performance. To overcome this, the present research work demonstrates an enhanced fraud detection system with Extreme Gradient Boosting (XGBoost) using cost-sensitive learning and SMOTE (Synthetic Minority Oversampling Technique). A baseline XGBoost classifier is first developed without class weighting. A cost-sensitive classifier is then developed using the scale_pos_weight parameter to increase the sensitivity of the minority class. SMOTE is then applied to further balance the training data to reduce false negatives. RandomizedSearchCV is employed for hyperparameter tuning. Classifiers are evaluated on precision, recall, F1-score, ROC-AUC, and the confusion matrix. Results indicate that SMOTE with cost-sensitive XGBoost significantly improves recall and AUC in identifying fraudulent transactions and is thus highly suitable for application in real-world financial fraud detection systems.

Item Type:	Thesis (Masters)
Supervisors:	Name Email Haque, Rejwanul UNSPECIFIED
Uncontrolled Keywords:	Credit card fraud detection; XGBoost; cost-sensitive learning; SMOTE; hyperparameter tuning; imbalanced data
Subjects:	H Social Sciences > HG Finance Q Science > QH Natural history > QH301 Biology > Methods of research. Technique. Experimental biology > Data processing. Bioinformatics > Artificial intelligence Q Science > Q Science (General) > Self-organizing systems. Conscious automata > Artificial intelligence H Social Sciences > HG Finance > Credit. Debt. Loans.
Divisions:	School of Computing > Master of Science in Artificial Intelligence for Business
Depositing User:	Ciara O'Brien
Date Deposited:	24 Aug 2026 10:59
Last Modified:	24 Aug 2026 10:59
URI:	https://norma.ncirl.ie/id/eprint/9592

Actions (login required)

View Item