Expanded Financial Fraud Data

Financial fraud dataset, originated from a collection of smaller datasets fused together and expanded with synthetic data, fuzzy filling, and noise-balanced filling.

The dataset contains up to 7 million entries, and is fully anonimized. Sensitive information is encoded as Float32 values, and the rest is kept as original, including decimal booleans, and Float32 values.

本数据集为金融欺诈数据集，由多个小型数据集融合得到，并通过合成数据、模糊填充与噪声均衡填充完成扩展。
该数据集包含至多700万条数据条目，且已完成全匿名化处理。敏感信息被编码为32位浮点数（Float32），其余字段保留原始格式，其中包括十进制布尔值与32位浮点数。