English (United Kingdom)

https://curated-unify.zendy.io/wp-json/zendy-region/v1/featured_content/oa?rat=en

https://curated-unify.zendy.io/wp-json/zendy-region/v1/highlighted_journal/

Global: Zendy Tools annual

Presents the keyphrase highlighting as premium feature

Keyphrase Highlighting

Presents the summarisation as premium feature

Summarisation

Insights

Presents the pdf analysis as premium feature

PDF Analysis

Presents the zaia usage as premium feature

ZAIA

Global: Zendy Tools monthly

Global: Zendy Plus monthly

Presents the access of premium content as premium feature

Premium Content

Global: Zendy Plus annual

Global: Zendy Free Plan

Deep neural networks (DNNs) have been broadly adopted in health riskprediction to provide healthcare diagnoses and treatments. To evaluate theirrobustness, existing research conducts adversarial attacks in thewhite/gray-box setting where model parameters are accessible. However, a morerealistic black-box adversarial attack is ignored even though most real-worldmodels are trained with private data and released as black-box services on thecloud. To fill this gap, we propose the first black-box adversarial attackmethod against health risk prediction models named MedAttacker to investigatetheir vulnerability. MedAttacker addresses the challenges brought by EHR datavia two steps: hierarchical position selection which selects the attackedpositions in a reinforcement learning (RL) framework and substitute selectionwhich identifies substitute with a score-based principle. Particularly, byconsidering the temporal context inside EHRs, it initializes its RL positionselection policy by using the contribution score of each visit and the saliencyscore of each code, which can be well integrated with the deterministicsubstitute selection process decided by the score changes. In experiments,MedAttacker consistently achieves the highest average success rate and evenoutperforms a recent white-box EHR adversarial attack technique in certaincases when attacking three advanced health risk prediction models in theblack-box setting across multiple real-world datasets. In addition, based onthe experiment results we include a discussion on defending EHR adversarialattacks.

MedAttacker: Exploring Black-Box Adversarial Attacks on Risk Prediction Models in Healthcare