Greedy Attack and Gumbel Attack: Generating Adversarial Examples for Discrete Data

Puyudi Yang
Puyudi Yang
Jianbo Chen
Jianbo Chen
Jane-Ling Wang
Jane-Ling Wang

Journal of Machine Learning Research, 2020.

Cited by: 19|Bibtex|Views21|Links
EI

Abstract:

We present a probabilistic framework for studying adversarial attacks on discrete data. Based on this framework, we derive a perturbation-based method, Greedy Attack, and a scalable learning-based method, Gumbel Attack, that illustrate various tradeoffs in the design of attacks. We demonstrate the effectiveness of these methods using bo...More

Code:

Data:

Your rating :
0

 

Tags
Comments