Scholarlink

Credit card fraud poses a significant challenge in financial systems, requiring advanced techniques for accurate detection. This chapter proposes a hybrid machine learning model that combines XGBoost and logistic regression to enhance fraud detection accuracy and recall. The proposed model achieves significant improvements over standalone models, particularly in handling imbalanced datasets. Additionally, we discuss future work in implementing federated learning frameworks to ensure secure, distributed fraud detection. Our results demonstrate a recall rate of 98% and a precision of 95%, validating the hybrid model’s efficacy in real-world scenarios.

错误：搜索内容不能为空，请输入英文关键词
错误：关键词超出字数限制，请精简
高级检索

Enhancing Credit Card Fraud Detection Using Hybrid Machine Learning Models with Federated Learning Prospects

  • 会议论文

  • 发表日期：

    2026年

查看全文
  • Raghed Mohamed,
  • Ahmad Haddad,
  • Eisa Alserkal,
  • Salem Alfalasi,
  • Alavikunhu Panthakkan,
  • Shadi Atalla

摘要

Credit card fraud poses a significant challenge in financial systems, requiring advanced techniques for accurate detection. This chapter proposes a hybrid machine learning model that combines XGBoost and logistic regression to enhance fraud detection accuracy and recall. The proposed model achieves significant improvements over standalone models, particularly in handling imbalanced datasets. Additionally, we discuss future work in implementing federated learning frameworks to ensure secure, distributed fraud detection. Our results demonstrate a recall rate of 98% and a precision of 95%, validating the hybrid model’s efficacy in real-world scenarios.