PHISHING AND FRAUD DETECTION IN IMBALANCED DATASETS: A COMPARATIVE STUDY OF ENSEMBLE AND DEEP LEARNING MODELS
Date
2026-05-12
Authors
Maria Khalil
Nael Abu Halaweh
Journal Title
Journal ISSN
Volume Title
Publisher
Deanship of Scientific Research - Al-Quds University
Abstract
Phishing and fraud attacks remain major challenges in cybersecurity, particularly due to class imbalance, where malicious cases are significantly fewer than legitimate ones. This imbalance complicates detection and often reduces model performance. Although deep learning is widely regarded as a powerful approach, it does not always perform effectively on tabular and highly imbalanced data, highlighting the need to evaluate alternative methods.