Towards One Shot Search Space Poisoning in Neural Architecture Search
Machine Learning
2021-11-16 v1 Artificial Intelligence
Neural and Evolutionary Computing
Abstract
We evaluate the robustness of a Neural Architecture Search (NAS) algorithm known as Efficient NAS (ENAS) against data agnostic poisoning attacks on the original search space with carefully designed ineffective operations. We empirically demonstrate how our one shot search space poisoning approach exploits design flaws in the ENAS controller to degrade predictive performance on classification tasks. With just two poisoning operations injected into the search space, we inflate prediction error rates for child networks upto 90% on the CIFAR-10 dataset.
Keywords
Cite
@article{arxiv.2111.07138,
title = {Towards One Shot Search Space Poisoning in Neural Architecture Search},
author = {Nayan Saxena and Robert Wu and Rohan Jain},
journal= {arXiv preprint arXiv:2111.07138},
year = {2021}
}
Comments
(Student Abstract) In Proceedings of the 36th AAAI Conference on Artificial Intelligence, Vancouver, BC,Canada, 2022. arXiv admin note: substantial text overlap with arXiv:2106.14406