Scholarlink

In OPFD, sampling methods can, to some extent, enhance the feature representation of fraudulent samples, but synthetic samples often fail to fully capture the dynamically evolving temporal patterns present in real-world transaction scenarios [19].

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

Learning Fraud-Sensitive Transactional Representations via Attention and Temporal Modeling

  • 章节

  • 发表日期：

    2026年

查看全文
  • Yu Xie,
  • Yue Tian,
  • Jiamin Yao,
  • Guanjun Liu

摘要

In OPFD, sampling methods can, to some extent, enhance the feature representation of fraudulent samples, but synthetic samples often fail to fully capture the dynamically evolving temporal patterns present in real-world transaction scenarios [19].