English

Indiscriminate Data Poisoning Attacks on Neural Networks

Machine Learning 2024-02-16 v2 Cryptography and Security

Abstract

Data poisoning attacks, in which a malicious adversary aims to influence a model by injecting "poisoned" data into the training process, have attracted significant recent attention. In this work, we take a closer look at existing poisoning attacks and connect them with old and new algorithms for solving sequential Stackelberg games. By choosing an appropriate loss function for the attacker and optimizing with algorithms that exploit second-order information, we design poisoning attacks that are effective on neural networks. We present efficient implementations that exploit modern auto-differentiation packages and allow simultaneous and coordinated generation of tens of thousands of poisoned points, in contrast to existing methods that generate poisoned points one by one. We further perform extensive experiments that empirically explore the effect of data poisoning attacks on deep neural networks.

Keywords

Cite

@article{arxiv.2204.09092,
  title  = {Indiscriminate Data Poisoning Attacks on Neural Networks},
  author = {Yiwei Lu and Gautam Kamath and Yaoliang Yu},
  journal= {arXiv preprint arXiv:2204.09092},
  year   = {2024}
}

Comments

Accepted to TMLR in 2022

R2 v1 2026-06-24T10:52:32.840Z