Subgraph Centralization: A Necessary Step for Graph Anomaly Detection

GCAD

This project implements the Subgraph Centralization: A Necessary Step for Graph Anomaly Detection for graph anomaly detection

Model Usage

Dependencies

numpy
scipy
scikit-learn

Input Format

Get start

Run the code with the following command.

python run.py --psi 2 --dataset cora --h 1 --lamda 0.0625
python run.py --psi 2 --dataset citeseer --h 1 --lamda 0.125
python run.py --psi 2 --dataset pubmed --h 1 --lamda 0.0625
python run.py --psi 2 --dataset ACM --h 1 --lamda 0.0625

python run.py --psi 2 --dataset lattice --h 1 --lamda 0.0
python run.py --psi 8 --dataset lattice_s --h 1 --lamda 0.003125
python run.py --psi 4 --dataset RGG_1 --h 2 --lamda 0.0
python run.py --psi 8 --dataset RGG_22 --h 1 --lamda 0.0
python run.py --psi 2 --dataset SBM_str --h 1 --lamda 0.25
python run.py --psi 2 --dataset watts_strogatz --h 1 --lamda 0.0