Defending Graph Neural Networks Against Adversarial Attacks via Matrix Factorization

This is the complete code for the RobustMF framework to defense Graph Neural Netowrks GNN against adverisrial Attacks. The work is pubished in the following papers:
(Paper)

Will be posted after it gets published 

Comments/Bugs/Problems:
maburidi@ucmerced.edu
maburidi@gmail.com

May, 2024. Initial release

Installation

First install dependencies:

!bash install_dependencies.sh

Also, install DeepRobust

!git clone https://github.com/DSE-MSU/DeepRobust.git
%cd DeepRobust
!python setup.py install

Train and test

Use Pertrubation rate = 10% , on Cora and using MetaAttach to perturb

!python main.py --ptb_rate 0.1 --dataset 'cora' --attack 'meta' --epochs 250  

Main Results:

Cite as:

Will be poseted once published