Skip to content

Repository of the paper "Imperceptible Adversarial Attacks on Tabular Data" presented at NeurIPS 2019 Workshop on Robust AI in Financial Services (Robust AI in FS 2019)

License

Notifications You must be signed in to change notification settings

axa-rev-research/LowProFool

Folders and files

NameName
Last commit message
Last commit date

Latest commit

 

History

7 Commits
 
 
 
 
 
 
 
 
 
 
 
 
 
 

Repository files navigation

Disclaimer: This repository is not maintained anymore


LowProFool

LowProFool is an algorithm that generates imperceptible adversarial examples

This GitHub hosts the code to replicate the experiments presented in the paper:

Ballet, V., Renard, X., Aigrain, J., Laugel, T., Frossard, P., & Detyniecki, M. (2019). Imperceptible Adversarial Attacks on Tabular Data. NeurIPS 2019 Workshop on Robust AI in Financial Services: Data, Fairness, Explainability, Trustworthiness, and Privacy (Robust AI in FS 2019)

https://arxiv.org/abs/1911.03274

Adverse.py

Contains the implementation of LowProFool [1] along with an modifier version of DeepFool [2] that handles tabular datasets.

Metrics.py

Implements metrics introduced in [1]

Playground.ipynb

A demo python notebook to generate adversarial examples on the German Credit dataset and compare the results to DeepFool

References

[1] Ballet, V., Renard, X., Aigrain, J., Laugel, T., Frossard, P., & Detyniecki, M. (2019). Imperceptible Adversarial Attacks on Tabular Data. NeurIPS 2019 Workshop on Robust AI in Financial Services: Data, Fairness, Explainability, Trustworthiness, and Privacy (Robust AI in FS 2019)

[2] S. Moosavi-Dezfooli, A. Fawzi, P. Frossard: DeepFool: a simple and accurate method to fool deep neural networks. In Computer Vision and Pattern Recognition (CVPR ’16), IEEE, 2016.

About

Repository of the paper "Imperceptible Adversarial Attacks on Tabular Data" presented at NeurIPS 2019 Workshop on Robust AI in Financial Services (Robust AI in FS 2019)

Resources

License

Stars

Watchers

Forks

Releases

No releases published

Packages

No packages published