Conference proceeding
MedAttacker: Exploring Black-Box Adversarial Attacks on Risk Prediction Models in Healthcare
2022 IEEE International Conference on Bioinformatics and Biomedicine (BIBM), pp.1777-1780
12/06/2022
DOI: 10.1109/BIBM55620.2022.9994898
Abstract
Researchers have conduct adversarial attacks against deep neural networks (DNNs) for health risk prediction in the white/gray-box setting to evaluate their robustness. However, since most real-world solutions are trained by private data and released as black-box services on the cloud, we should investigate their robustness in the black-box setting. Unfortunately, existing work ignores to consider the uniqueness of electronic health records (EHRs). To fill this gap, we propose the first black-box adversarial attack method against health risk prediction models named MedAttacker to investigate their vulnerability. It addresses the challenges brought by EHRs via two steps: hierarchical position selection which selects the attacked positions in a reinforcement learning (RL) framework and substitute selection which identifies substitutes with a score-based principle. Particularly, by considering the temporal context inside EHRs, MedAttacker initializes its RL position selection policy by using the contribution score of each visit and the saliency score of each code, which can be well integrated with the deterministic substitute selection process decided by the score changes. We evaluate MedAttacker by attacking three advanced risk prediction models in the black-box setting across multiple real-world datasets, and MedAttacker consistently achieves the highest average success rate and even outperforms a recent white-box EHR adversarial attack technique in certain cases.
Details
- Title: Subtitle
- MedAttacker: Exploring Black-Box Adversarial Attacks on Risk Prediction Models in Healthcare
- Creators
- Muchao Ye - Pennsylvania State UniversityJunyu Luo - Pennsylvania State UniversityGuanjie Zheng - Shanghai Jiao Tong UniversityCao Xiao - Relativity,USAHouping Xiao - Georgia State UniversityTing Wang - Pennsylvania State UniversityFenglong Ma - Pennsylvania State University
- Resource Type
- Conference proceeding
- Publication Details
- 2022 IEEE International Conference on Bioinformatics and Biomedicine (BIBM), pp.1777-1780
- Publisher
- IEEE
- DOI
- 10.1109/BIBM55620.2022.9994898
- Grant note
- National Science Foundation (10.13039/100000001)
- Language
- English
- Date published
- 12/06/2022
- Academic Unit
- Computer Science
- Record Identifier
- 9984696574802771
Metrics
1 Record Views