Skip to content
/ EpoAtk Public

The implementation of our ICDM 2020 paper "Exploratory Adversarial Attacks on Graph Neural Networks".

Notifications You must be signed in to change notification settings

EpoAtk/EpoAtk

Folders and files

NameName
Last commit message
Last commit date

Latest commit

 

History

10 Commits
 
 
 
 

Repository files navigation

EpoAtk

The implementation of our ICDM 2020 paper "Exploratory Adversarial Attacks on Graph Neural Networks".

Requirements

python == 3.6.8
torch == 1.1.0
numpy == 1.16
scipy == 1.2.1
networkx == 1.11

Preprocess

unzip data.zip in ./src.

Attack

cd ./src.
zsh scripts with default hyper-parameters.

Citation

If you find the code is useful for your research, please cite this paper:

@inproceedings{lin2020:epo,
author={Lin, Xixun and Zhou, Chuan and Yang, Hong and Wu, Jia and Wang, Haibo and Cao, Yanan and Wang, Bin},
title={Exploratory Adversarial Attacks on Graph Neural Networks},
booktitle={IEEE International Conference On Data Mining (ICDM)},
year={2020}
}

About

The implementation of our ICDM 2020 paper "Exploratory Adversarial Attacks on Graph Neural Networks".

Topics

Resources

Stars

Watchers

Forks

Releases

No releases published

Packages

No packages published