Communications of the ACM
NIST Promotes Testbed to Address Threats Targeting ML Systems
The testbed follows NIST Internal Report 8269, a taxonomy of adversarial machine learning that the agency published in 2019 which identifies three major categories of attacks against ML algorithms: evasion, poisoning, and oracle.
Credit: Getty Images
The National Institute of Standards and Technology's (NIST) National Cybersecurity Center of Excellence has demonstrated the first iteration of an experimental testbed to address cybersecurity threats targeting machine learning (ML) algorithms.
The testbed, called Dioptra, will enable researchers to assess security techniques and solutions for safeguarding ML-enabled systems, testing various combinations of attacks, defenses, and model architectures.
Said NIST computer scientist Harold Booth, "The basic take-home message is that there are a lot of metrics and measurements out there, but you really want to give some thought to what [solutions] you're using and make sure you understand how those play with your deployments."
From FCW
View Full Article
Abstracts Copyright © 2022 SmithBucklin, Washington, DC, USA
No entries found