Quantum adversarial machine learning
135 Citations•2020•
Sirui Lu, Lu-Ming Duan, Dong-Ling Deng
The results uncover the notable vulnerability of quantum machine learning systems to adversarial perturbations, which not only reveals a novel perspective in bridging machine learning and quantum physics in theory but also provides valuable guidance for practical applications of quantum classifiers based on both near-term and future quantum technologies.
Abstract
This work uncovers the vulnerability aspect for quantum machine learning, by showing that quantum classifiers are vulnerable to adversarial perturbations. The authors give generic recipes on how to generate adversarial perturbations and mitigate the vulnerability problem in various adversarial scenarios.