Odyssey: A Systems Approach to Machine Learning Security

Abstract

This paper provides a systems approach to addressing attacks, consequences, and mitigations for systems using Machine Learning (ML). It explains each of these over the lifecycle of an ML technology, providing clear explanations of what to worry about, when to worry about it, and how to mitigate it while presuming little incoming knowledge of ML specifics. Our discussion of ML vulnerabilities, attacks, and mitigations utilizes the taxonomy developed in NISTIR 8269.

Open PDF

Document Details

Document Type
Technical Report
Publication Date
Apr 01, 2021
Accession Number
AD1157105

Entities

People

  • Chris Giannella
  • Jones Malachi
  • Joseph Jubinski
  • Ransom Winder

Organizations

  • MITRE Corporation

Tags

Communities of Interest

  • Autonomy

DTIC Thesaurus Topics

  • Addressing
  • Artificial Intelligence
  • Artificial Intelligence Software
  • Computers
  • Computing Devices
  • Cybersecurity
  • Data Mining
  • Dimensionality Reduction
  • Information Processing
  • Information Science
  • Information Systems
  • Learning
  • Machine Learning
  • Neural Networks
  • Reinforcement Learning
  • Security
  • Supervised Machine Learning
  • Systems Approach
  • Taxonomy
  • Unsupervised Machine Learning
  • Vulnerability

Fields of Study

  • Computer science

Readers

  • Agent-Based Social Robotics and Mobile-Assisted Learning in Virtual Environments.
  • Cybersecurity.

Technology Areas

  • AI & ML
  • AI & ML - DoD AI Strategy
  • AI & ML - Neural Networks