BOND
收藏官方服务：
资源简介：
用于静态属性图上无监督异常节点检测的第一个综合基准，评估了14种方法的异常检测性能，并使用九个真实数据集来评估不同检测方法对两种主要类型的合成异常和真实非合成异常的响应。
The first comprehensive benchmark for unsupervised anomalous node detection on static attributed graphs. It evaluates the anomaly detection performance of 14 methods, and employs nine real-world datasets to assess the responses of different detection methods to two main types of synthetic anomalies and real non-synthetic anomalies.
创建时间：
2022-06-21
