Skip to content
/ ADA-GAD Public

Official PyTorch implementation for the paper ADA-GAD: Anomaly-Denoised Autoencoders for Graph Anomaly Detection (AAAI 2024).

License

Notifications You must be signed in to change notification settings

jweihe/ADA-GAD

Repository files navigation

This repository contains a PyTorch implementation for our paper "ADA-GAD: Anomaly-Denoised Autoencoders for Graph Anomaly Detection".

Framework Details

Framework Details

We introduce a novel twostage framework called Anomaly-Denoised Autoencoders for Graph Anomaly Detection (ADA-GAD). In the first stage, we design a learning-free anomaly-denoised augmentation method to generate graphs with reduced anomaly levels. We pretrain graph autoencoders on these augmented graphs at multiple levels, which enables the graph autoencoders to capture normal patterns. In the next stage, the decoders are retrained for detection on the original graph, benefiting from the multi-level representations learned in the previous stage.

Getting Started

Environment Setup

git clone https://github.com/jweihe/ADA-GAD.git
cd ADA-GAD
conda create -n ada-gad python=3.8
conda activate ada-gad
pip install -r requirements.txt

Main Usage

Use the following command to run the main script with configuration options:

python main.py --use_cfg --seeds 0 --dataset $dataset

We support using datasets from data. You can download these datasets and place them in the data directory data. Alternatively, you can run the code directly, and the dataset will be downloaded automatically.

Citation

@article{he2023ada,
  title={ADA-GAD: Anomaly-Denoised Autoencoders for Graph Anomaly Detection},
  author={He, Junwei and Xu, Qianqian and Jiang, Yangbangyan and Wang, Zitai and Huang, Qingming},
  journal={arXiv preprint arXiv:2312.14535},
  year={2023}
}

Acknowledgements

This code is built on pygod and GraphMAE. We thank the authors for sharing the codes.

About

Official PyTorch implementation for the paper ADA-GAD: Anomaly-Denoised Autoencoders for Graph Anomaly Detection (AAAI 2024).

Resources

License

Stars

Watchers

Forks

Releases

No releases published

Packages

No packages published