A PyTorch implementation of "[Few Sampled Anomalies are All You Need: A Framework for Enhanced Graph Anomaly Detection]"
Dependencies for all experiments of MGAD are as follows:
• Python == 3.7.8
• PyTorch == 1.6.0
• NetworkX == 2.6.3
• Scikit-learn == 0.23.2
• NumPy == 1.18.5
• SciPy == 1.7.3
All of datasets used in this paper are put in ./data folder and graph information (e.g., adjacency, attribute, and label) is included in each dataset file (.mat). Due to the large file size, some of metapaths are included in that folder as .csv files. Various metapath schema can be extracted by using different metapath extractors. The meaning of filename is {dataset} {metapath schema} {anomaly ratio} {sampling round}.csv.
python main.py --dataset cora