Deep Detector Health Management under Adversarial Campaigns

Javier Echauz, Keith Kenemer, Sarfaraz Hussein, Jay Dhaliwal, Saurabh Shintre, Slawomir Grzonkowski, and Andrew Gardner
Publication Target: 
Publication Issue: 
Special Issue on Deep Learning and Emerging Analytics
Submission Type: 
Full Paper
ijphm_19_032.pdf4.35 MBJanuary 3, 2020 - 9:42am

Machine learning models are vulnerable to adversarial inputs that induce seemingly unjustifiable errors. As automated classifiers are increasingly used in industrial control systems and machinery, these adversarial errors could grow to be a serious problem. Despite numerous studies over the past few years, the field of adversarial ML is still considered alchemy, with no practical unbroken defenses demonstrated to date, leaving PHM practitioners with few meaningful ways of addressing the problem. We introduce “turbidity detection” as a practical superset of the adversarial input detection problem, coping with adversarial campaigns rather than statistically invisible one-offs. This perspective is coupled with ROC-theoretic design guidance that prescribes an inexpensive domain adaptation layer at the output of a deep learning model during an attack campaign. The result aims to approximate the Bayes optimal mitigation that ameliorates the detection model’s degraded health. A “proactively reactive” type of prognostics is achieved via Monte Carlo simulation of various adversarial campaign scenarios, by sampling from the model’s own turbidity distribution to quickly deploy the correct mitigation during a real-world campaign.

Publication Year: 
Publication Volume: 
Publication Control Number: 
Page Count: 
Submission Keywords: 
deep convolution neural network
binary classifier
Asset health management
Submission Topic Areas: 
Automated reconfiguration
CBM and informed logistics
Data-driven methods for fault detection, diagnosis, and prognosis
Health management system design and engineering
Software health management
Submitted by: 

follow us

PHM Society on Facebook Follow PHM Society on Twitter PHM Society on LinkedIn PHM Society RSS News Feed