Error-Bounded Graph Anomaly Loss for GNNs
This repository contains the code package for the TNNLS paper:
A Synergistic Approach for Graph Anomaly Detection with Pattern Mining and Feature Learning.
Authors: Tong Zhao (tzhao2@nd.edu), Tianwen Jiang, Neil Shah, and Meng Jiang.
and the CIKM'20 paper:
Error-Bounded Graph Anomaly Loss for GNNs.
Authors: Tong Zhao (tzhao2@nd.edu), Chuchen Deng, Kaifeng Yu, Tianwen Jiang, Daheng Wang, and Meng Jiang.
Usage
1. Dependencies
This code package was developed with Python 3.6.8 and PyTorch 1.0.1.post2.
A detailed dependencies list can be found in requirements.txt and can be installed by:
pip install -r requirements.txt
2. Data
Data files are located at /data/[dataset]/, a simple example of loading the data can be found here. Specifically, [dataset]_graph_u2p.pkl is the pickled sparse adjacency matrix (csr_matrix) and [dataset]_labels_u.pkl is the pickled user labels.
3. Run
To train the model, run
python -m src.main
list of arguments can be found at here.
Cite
If you find this repository useful in your research, please cite our papers:
@ARTICLE{zhao2021synergistic,
author={Zhao, Tong and Jiang, Tianwen and Shah, Neil and Jiang, Meng},
journal={IEEE Transactions on Neural Networks and Learning Systems},
title={A Synergistic Approach for Graph Anomaly Detection With Pattern Mining and Feature Learning},
year={2021},
volume={},
number={},
pages={1-13},
doi={10.1109/TNNLS.2021.3102609}}
@inproceedings{zhao2020error,
title={Error-Bounded Graph Anomaly Loss for GNNs},
author={Zhao, Tong and Deng, Chuchen and Yu, Kaifeng and Jiang, Tianwen and Wang, Daheng and Jiang, Meng},
booktitle={Proceedings of the 29th ACM International Conference on Information \& Knowledge Management},
pages={1873--1882},
year={2020}
}