Accession Number:

AD1157105

Title:

Odyssey: A Systems Approach to Machine Learning Security

Descriptive Note:

[Technical Report, Technical Report]

Corporate Author:

MITRE CORP ANNAPOLIS JUNCTION MD

Personal Author(s):

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

Report Date:

2021-04-01

Pagination or Media Count:

43

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.

Descriptors:

• supervised machine learning
• neural networks
• artificial intelligence software
• data mining
• information science
• machine learning
• artificial intelligence
• cybersecurity
• unsupervised machine learning
• dimensionality reduction
• information processing
• information systems
• security
• systems approach
• computers
• computing devices
• reinforcement learning

Subject Categories:

• Cybernetics
• Administration and Management

Distribution Statement:

[A, Approved For Public Release]