AOF (Account Opening Fraud)
收藏arXiv2025-09-30 收录
下载链接：
https://github.com/feedzai/fair-automl
下载链接
链接失效反馈官方服务：
资源简介：
该数据集是一个大规模的银行账户开户申请案例研究，包含了超过50万条实例，这些数据源自6个月的数据流。该数据集用于验证在考虑到公平性的情况下进行超参数优化的有效性。此外，该数据集中的特征存在严重的类别不平衡问题，大约每99个负标签样本对应1个正标签样本。在规模上，这是一个大型数据集，包含超过50万条实例，其任务是对银行账户申请进行欺诈检测。
This is a large-scale dataset for case studies of bank account opening applications, containing over 500,000 instances sourced from 6 months of streaming data. It is utilized to validate the effectiveness of hyperparameter optimization that incorporates fairness considerations. Additionally, the dataset exhibits severe class imbalance, with approximately 99 negative-label samples corresponding to 1 positive-label sample. As a large-scale dataset with over 500,000 instances, its target task is fraud detection for bank account opening applications.
提供机构：
Private dataset related to bank account applications
