This repository contains the source code for the paper titled "Dual-channel Graph-level Anomaly Detection Method Based on Multi-Graph Representation Learning." The code implements our proposed method for detecting anomalies in graph data using multi-graph representation learning techniques.
├── data/
│ └── Datasets/ #Place your datasets here
├── model/
│ └── tracemgfr.py # Our Model
├── utils/
│ ├── functions.py # functions
The datasets used for experiments are located in the data/ directory. You can create subdirectories for different datasets, e.g., data/Datasets/, and place your dataset files there.
The model implementation can be found in the model/ directory. The main file, tracemgfr.py , contains the architecture and training routines for our dual-channel graph-level anomaly detection method. The other .py files are the code for the baseline model.
python model/tracemgfr.py
Figure 1: Overview of the dual-channel anomaly detection framework.
JING,Y., Chen,J., Chen,X.,Wang,H.: Dual-channel Graph-level Anomaly Detection Method Based on Multi-Graph Representation Learning. Applied Intelligence.(Contact:jyj.chn@gmail.com)