Designing Adversarially Resilient Classifiers using Resilient Feature Engineering

12/17/2018
by   Kevin Eykholt, et al.
0

We provide a methodology, resilient feature engineering, for creating adversarially resilient classifiers. According to existing work, adversarial attacks identify weakly correlated or non-predictive features learned by the classifier during training and design the adversarial noise to utilize these features. Therefore, highly predictive features should be used first during classification in order to determine the set of possible output labels. Our methodology focuses the problem of designing resilient classifiers into a problem of designing resilient feature extractors for these highly predictive features. We provide two theorems, which support our methodology. The Serial Composition Resilience and Parallel Composition Resilience theorems show that the output of adversarially resilient feature extractors can be combined to create an equally resilient classifier. Based on our theoretical results, we outline the design of an adversarially resilient classifier.

READ FULL TEXT

page 1

page 2

page 3

page 4

research
01/28/2022

Positive Rate Binary Interactive Error Correcting Codes Resilient to >1/2 Adversarial Erasures

An interactive error correcting code (𝗂𝖤𝖢𝖢) is an interactive protocol w...
research
05/21/2021

Resilience in the Cyber World: Definitions, Features and Models

Resilience is a feature that is gaining more and more attention in compu...
research
09/25/2021

Beyond Robustness: A Taxonomy of Approaches towards Resilient Multi-Robot Systems

Robustness is key to engineering, automation, and science as a whole. Ho...
research
07/03/2022

Graph Learning based Generative Design for Resilience of Interdependent Network Systems

Interconnected complex systems usually undergo disruptions due to intern...
research
12/05/2022

Resilience Evaluation of Entropy Regularized Logistic Networks with Probabilistic Cost

The demand for resilient logistics networks has increased because of rec...
research
04/14/2020

Resilient Self-Debugging Software Protection

Debuggers are a popular reverse engineering and tampering tool. Self-deb...

Please sign up or login with your details

Forgot password? Click here to reset