Implementations of ICLR 2024 submission "Deceptive Fairness Attacks on Graphs via Meta Learning"
Main dependency: python 3.8, pytorch 1.12.1, deeprobust 0.2.4
To generate the poisoned graph with FATE, go to src/ folder and run the following code:
python fate_attack.py --dateset pokec_n --fairness statistical_parity --ptb_mode flip --ptb_rate 0.05 --attack_steps 3 --attack_seed 25
To train the victim model, go to src/ folder and run the following code:
python train.py --dateset pokec_n --fairness statistical_parity --ptb_mode flip --ptb_rate 0.05 --attack_steps 3 --attack_seed 25 --attack_method fate --victim_model gcn --hidden_dimension 128 --num_epochs 400
To test under different settings, please feel free to refer to the detailed parameter settings listed in Appendix D.
Deceptive Fairness Attacks on Graphs via Meta Learning is licensed under CC BY-NC-ND 4.0. To view a copy of this license, visit http://creativecommons.org/licenses/by-nc-nd/4.0/