HC-GLAD: Dual Hyperbolic Contrastive Learning for Unsupervised Graph-Level Anomaly Detection

Requirements

This code requires the following dependencies:

• Python == 3.9.19
• PyTorch == 1.11.0
• PyTorch Geometric == 2.0.4
• NumPy == 1.26.4
• Scikit-learn == 1.0.2
• OGB == 1.3.3
• NetworkX == 2.7.1
• Einops == 0.8.0

Hardware Infrastructures

The hardware configuration used for the implementation on the Linux server includes the following:

• CPU: Intel(R) Xeon(R) Gold 5220 CPU @ 2.20GHz
• GPU: NVIDIA A40 (48GB)

Datasets

Download and process automatically from the TUDataset https://chrsmrrs.github.io/datasets/docs/datasets/

Quick Start

run the script corresponding to the dataset you want. For instance:

bash script/ad_BZR.sh